| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Medicine (K.V.W., D.E.K.) and Radiology (J.C.P.), University of Pittsburgh, Pittsburgh, Pennsylvania 15261; Department of Information Engineering, University of Padova (A.B., B.M., C.C.), 35131 Padova, Italy
Address all correspondence and requests for reprints to: Katherine Williams, M.D., M.P.H., Division of Endocrinology and Metabolism, University of Pittsburgh School of Medicine, 809N Montefiore University Hospital, 3459 Fifth Avenue, Pittsburgh, Pennsylvania 15213. E-mail: williamsk{at}msx.dept-med.pitt.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In prior studies, PET imaging with the glucose analog [18F]2-deoxy-2-D-fluoro-glucose (18F-FDG) has been found to be a highly sensitive method for the study of insulin action in skeletal muscle in health and IR (9, 11, 12, 13). The high specific activity of the tracer 18F-FDG permits in vivo human use of a deoxyglucose compound so that metabolic steps up to, but not beyond, glucose phosphorylation can be monitored (14). The steps that potentially influence 18F-FDG metabolism are 1) substrate delivery (tissue perfusion and capillary diffusion), 2) trans-membrane bidirectional transport, and 3) cellular trapping of glucose by phosphorylation. The challenge is to discern these three steps from the use of a single isotope. To analyze the dynamic data, our group has previously used the three-rate constant (3K; inward and outward transport, and phosphorylation) and three-compartment (plasma 18F-FDG, tissue 18F-FDG, and tissue 18F-FDG-6-phophate pools) model developed by Sokoloff (15) for use in the brain. A fundamental assumption of the original 3K model is that in cerebral tissue, transport of 18F-FDG across the cell membrane is very fast compared with its transport across the blood-brain barrier and compared with phosphorylation. Therefore, the concentration ratio between the interstitial space and the cellular space is nearly the same at all times. However, because of the structural and functional dissimilarities between brain and muscle, this assumption might not be applicable to skeletal muscle, where glucose enters via diffusion through the capillary wall.
The three rate constants of the 3K model (K1', k2', and k3') pertain to inward transport, outward transport, and phosphorylation, respectively. In our prior studies we have found only a minor effect of insulin on values of K1' even among insulin-sensitive subjects (13) and those with insulin resistance (16). This raises concern about whether this parameter portrays information on glucose transporter-mediated trans-membrane flux of glucose or is instead a more global index of the transfer of tracer from plasma to tissue and thus more closely reflects the dynamics of blood flow, tissue perfusion, and interstitial delivery. Also, we observed a robust insulin stimulation of the parameter k3' in insulin-sensitive individuals and a marked blunting of this in IR (9, 13, 16). This is the rate constant to which the kinetics of glucose phosphorylation is attributed, and therefore, these findings have been one basis for the assertion that this step of glucose metabolism is an important locus over insulin action in skeletal muscle.
Recently, a new four-compartment [plasma, extracellular, tissue 18F-FDG, and 18F-FDG-6-phosphate (18F-FDG-6-P)], five-rate constant (5K; K1, k2 plasma-extracellular exchange; k3, k4 transport in and out of cell; k5 phosphorylation) model was proposed by Bertoldo and applied to lean subjects (17). In the initial application of the 5K model, using dynamic PET data from lean, insulin-sensitive individuals, insulin was found to increase both the transport and the phosphorylation parameters, whereas the rate constant for exchange from plasma to the extracellular space was unchanged (17). In the current study we used the novel model to study insulin resistance across a range of doses of insulin stimulation in obesity and have compared the findings with those obtained using the classic 3K model and with data from lean healthy volunteers.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Subjects
Twelve lean and 13 obese, glucose-tolerant subjects were recruited by advertisement and randomly assigned to euglycemic insulin infusion studies at rates of 0 (n = 4 lean and 4 obese), 40 (n = 4 lean and 4 obese), and 120 (n = 4 lean and 5 obese) mU/m2·min. The study design is shown in Fig. 1
. The groups of lean and obese subjects were matched for age (37 ± 1 and 42 ± 2 yr), but differed for body mass index (24.9 ± 0.4 and 32.0 ± 0.6 kg/m2; P < 0.01). Each group was comprised predominately of male volunteers (10 men and 2 women, and 10 men and 3 women for the lean and obese groups, respectively). Before participating in this study, each subject had a medical examination to ensure that the subject was in good general health. Informed consent was obtained from each volunteer, and the investigation was reviewed and approved by the University of Pittsburgh institutional review board.
|
Subjects were admitted to the University of Pittsburgh General Clinical Research Center on the evening before the studies and fasted overnight after dinner. In the morning an iv catheter for infusion of insulin and glucose and for injection of 18F-FDG, and a radial artery catheter were placed, the latter for determination of 18F-FDG in plasma to be used as an input function in modeling. Before PET imaging, steady state metabolic conditions were attained using the glucose clamp procedure (18) for 3 h. Euglycemic insulin infusion was maintained during PET imaging. PET imaging studies of 18F-FDG uptake into midthigh skeletal muscle were performed at the University of Pittsburgh Positron Emission Tomography Center. Subjects were positioned in the PET scanner so that the midthigh corresponded to the midpoint axial field of view. After a transmission scan, 4 mCi 18F-FDG were injected iv over 30 sec. A 90-min dynamic PET scan was simultaneously initiated (19 frames: 4 for 30 sec, 4 for 2 min, 6 for 5 min, 5 for 10 min). The PET scans were acquired in 2- and 3-dimensional (2D and 3D) imaging modes using a Siemens (New York, NY) CTI 951 R/31 (lean, n = 5; obese, n = 6) scanner and an ECAT ART scanner (lean, n = 7; obese, n = 7), respectively. The imaging characteristics of the 2 scanners were comparable. The Siemens 951R/31 scanner acquired 31 imaging planes simultaneously [2D; in-plane resolution, 6.0 mm FWHM (ramp filter); axial slice width, 3.4 mm], whereas 47 imaging planes were acquired using the ECAT ART scanner [3D; in-plane resolution, 6.0 mm FWHM (ramp filter); axial slice width, 3.4 mm]. The scatter fraction was low for the 2D Siemens CTI 951 (13%) (19), and no scatter correction was performed following conventional methods. The 3D ART had a scatter fraction that was approximately 37% (20), and these emission data were corrected for scattered photons using a model-based correction method (21). After correction of the PET data for radioactive decay, the tissue time activity data were converted to units of radioactivity concentration (microcuries per milliliter) using an empiric phantom-based calibration factor (microcuries per milliliter per PET counts per pixel).
Sampling of arterial blood for plasma 18F-FDG radioactivity began simultaneously with PET scanning. Arterial samples were obtained at 6-sec intervals for 2 min; at 20-sec intervals for 1 min; at 30-sec intervals for 1 min; at 5, 7, 10, 15, 20, and 30 min; and then every 15 min until 90 min postinjection of 18F-FDG. The exact timing of each sample was recorded. Blood was centrifuged, and 200 µl plasma were removed for assay of plasma radioactivity using a Canbarra well counter (Packard, Downers Grove, IL). The counts per minute value for each sample was corrected for radioactive decay and converted to units of microcuries per milliliter based on the well counter sensitivity.
As previously described (13), to clearly define skeletal muscle on PET images, three cross-sectional computed tomography scans of 1-cm thickness were obtained at upper, mid, and lower boundaries of the midthigh and were visually coregistered with the matching PET transmission images. Regions of interest were drawn on medial, posterior, and lateral thigh muscles and applied to the dynamic PET scans across multiple (n = 20–30) planes. Tissue activities of 18F-FDG were expressed as microcuries per milliliter.
Modeling of 18F-FDG PET data
The time-activity curves of 18F-FDG images in skeletal muscle were first analyzed using the spectral analysis method, as previously described (22). Spectral analysis allows characterization of the reversible and irreversible components of the system and estimation of the minimum number of compartments needed to describe the data. Results (not shown) were very similar to those presented previously (17) and supported the use of the four-compartment, five-rate constant model. The model is shown in Fig. 2 and can be viewed as an extension of the classic 3K model reported by Sokoloff et al. (15) (Fig. 3), originally proposed in the brain. The novelty of the 5K model lies in an explicit delineation of an extracellular compartment, i.e. it distinguishes the kinetics steps of delivery of 18F-FDG to the extracellular space, its transport from the extracellular into the intracellular space, and its intracellular phosphorylation. The 5K model is described by:
 |
 |
 |
|
|
 |
 |
.
All four model parameters, K1', k2', k3', and Vb', are a priori identifiable (23). The fractional uptake of 18F-FDG, K' (milliliters per milliliter per minute), is calculated as:
 |
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the absence of insulin infusion (i.e. 0 mU/m2·min), obese volunteers had moderately higher fasting values for plasma insulin (53 ± 12 pmol in lean and 87 ± 13 pmol in obese; P = 0.06) and glucose (89 ± 1 mg/dl in lean; 99 ± 4 mg/dl in obese; P < 0.05) compared with lean subjects, and by study design no glucose was systemically infused. At the insulin infusion rate of 40 mU/m2·min, clamped values for plasma glucose were closely comparable in the two groups (94 ± 4 mg/dl in lean and 93 ± 2 mg/dl in obese). The rate of glucose infusion was substantially greater in lean subjects (7.7 ± 0.7 mg/kg·min in lean and 2.9 ± 0.3 mg/kg·min in obese; P < 0.01). At the insulin infusion rate of 120 mU/m2·min, the mean rate of systemic glucose infusion was similar in lean and obese subjects (8.3 ± 0.4 mg/kg·min in lean and 7.3 ± 1.2 mg/kg·min in obese), and plasma glucose values were nearly identical (92 ± 3 mg/dl in lean and 93 ± 2 mg/dl in obese). Steady state plasma insulin concentrations at both 40 and the 120 mU/m2·min were higher in obese volunteers (40 mU/m2·min: 446 ± 17 pmol in lean and 581 ± 46 pmol in obese; 120 mU/m2·min: 1271 ± 64 pmol in lean and 1979 ± 154 pmol in obese mU/m2·min; both P < 0.05), consistent with lesser insulin clearance in obesity.
5K model: transport and phosphorylation
The tissue time-activity data in lean and obese volunteers are shown in Fig. 4. Results are shown in Table 1 and Fig. 5.
|
|
|
Obese volunteers. In obese subjects, moderate and high rates of insulin infusion did not significantly change basal values for K1 or k2. Overall, values for K1 and k2 were quite comparable in lean and obese groups across the three rates of insulin infusion. In obese volunteers, the parameter k3, the tissue transport parameter, did not increase significantly above basal values in response to the moderate rate of insulin infusion, but did increase significantly in response to the high rate of insulin infusion (P < 0.05). There were no differences between lean and obese subjects for values of k3 during basal conditions, but at both levels of insulin stimulation, the obese volunteers had significantly lower values for k3. In obese volunteers, there was not a significant effect of insulin infusion on values for k4, nor were there differences between lean and obese volunteers in this parameter. In obesity, the phosphorylation parameter, k5, doubled in value during insulin-stimulated compared with basal conditions, but this change was not statistically significant (P = 0.32). In a similar manner, k5 was almost 50% lower in obese subjects compared with lean subjects during moderate insulin-stimulated conditions, but this difference did not reach statistical significance. Insulin stimulated 8-fold (40 mU/min·m2) and 24-fold (120 mU/min·m2) increases in fractional 18F-FDG uptake, K, from its basal value. K values were lower for obese than lean subjects in the basal state (P = 0.08) and during a 40 mU/min·m2 insulin stimulation (P < 0.05), whereas there were no differences between lean and obese volunteers in this parameter during the 120 mU/min·m2 insulin infusion.
5K model: [18F]FDG mass ratio and control coefficients
The results for the ratio of 18F-FDG masses in the intra- and extracellular spaces, Mint/Mext, as determined by the 5K model are shown in Fig. 6. The ratio was similar in lean and obese subjects during basal conditions (0.81 ± 0.14 vs. 0.64 ± 0.04 for lean and obese, respectively). In lean subjects the moderate rate of insulin infusion increased this ratio by 6-fold (from 0.81 ± 0.14 to 5.85 ± 2.49), which was a substantially greater than the 2-fold increase that occurred in obese subjects (from 0.64 ± 0.037 to 1.21 ± 0.30). However, the ratio Mint/Mext was equivalent in the two groups at the high rate of insulin infusion (8.83 ± 1.35 and 7.71 ± 3.25 for lean and obese, respectively).
|
. During basal conditions, phosphorylation contributes 69% of the control of glucose metabolism in lean subjects, with 31% residing with the transport coefficient. This situation was nearly identical in obese subjects for whom the phosphorylation coefficient contributes 66%, with 34% from the transport coefficient. However, during moderate insulin stimulation the distribution of control changed, and there were group differences. In lean volunteers during moderate insulin stimulation, 71% of the control over 18F-FDG uptake resided within transmembrane transport, and this did not change further during high insulin stimulation. In obese subjects during moderate insulin stimulation, the distribution of control was approximately 55% transmembrane transport and 45% phosphorylation. During high insulin stimulation in obese volunteers, the distribution further changed and matched that in lean volunteers, with 74% of control residing at transmembrane transport.
|
Although the 3K and 5K models differ in structure, both seek to describe the metabolism of 18F-FDG within tissue; accordingly, we sought to compare the results of these two models. Results from the 3K model are shown in Table 2. Regression analysis was used to compare rate constants from 3K and 5K models in each subject. There was a strong correlation between K1' and K1 (r = 0.88; P < 0.05), yet K1' did not have a significant correlation with k3. This suggests that in skeletal muscle the K1' parameter obtained from the 3K model is strongly shaped by the kinetics of tracer transfer from plasma to interstitial space rather than by the kinetics of 18F-FDG transport across the sarcolemma. The k3' parameter of the 3K model was significantly associated with both k3 (r = 0.78; P < 0.05) and k5 (r = 0.53; P < 0.05), findings suggesting that the k3' parameter is probably a composite parameter incorporating kinetics of both 18F-FDG transport and its phosphorylation. The fractional 18F-FDG uptake values of the 3K model are quite similar to those of the 5K model.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The 5K model for analysis of dynamic 18F-FDG data in skeletal muscle used in the current study was recently developed and described by Bertoldo et al. (17). The unique feature of the 5K model, one that adapts it specifically for skeletal muscle, is that it contains a compartment for the movement of tracer from plasma to extracellular space, or interstitial space within skeletal muscle, and two additional rate constants to describe the kinetics of exchange with this compartment and plasma. The classic three-compartment, three-rate constant model, as developed by Sokoloff et al. (15) for use with 18F-FDG in the study of central nervous system glucose metabolism, did not have such a compartment.
The rationale to add an extracellular compartment to the classic configuration of a plasma compartment, an intracellular free 18F-FDG, and an intracellular 18F-FDG-6-P compartment stems from two key considerations. One is anatomical. It is well known that skeletal muscle contains a substantially larger interstitial space than does brain (
50% greater in volume) and that interstitial space of muscle contains a greater amount of collagen (25). Moreover, as at rest only approximately one quarter of muscle capillaries are open (25), diffusion distances may vary more in muscle than in brain. These characteristics are in contrast to those of the central nervous system, which has a smaller interstitial volume, and therefore, it is with greater accuracy that it can be assumed, as with the classic model, that tracer moves rapidly from plasma to the cell surface. Given the larger interstitial space of skeletal muscle compared with brain, another consideration is that glucose metabolism in muscle, unlike that in brain, can be strongly affected by insulin. Leg balance studies reveal that arterio-venous fractional extraction of glucose during insulin-stimulated conditions often exceed 20% and can approach 50% in insulin-sensitive individuals, while being quite small, barely greater than the 1–3% of basal conditions in insulin-resistant individuals (13). Given the large arterio-venous differences that can exist, it follows that a concomitantly large gradient would exist within the diffusion radius of interstitial space from capillary to tissue. It has been previously suggested that interstitial gradients for glucose can be a determinant of rates of insulin-stimulated tissue utilization (26).
The second key consideration for including an extracellular compartment derives from the results of a relatively model-independent analysis of the dynamic PET 18F-FDG data in skeletal muscle of healthy volunteers based on spectral analysis (22, 27), which suggested the existence of two reversible and one irreversible compartment. Spectral analysis of the data from lean and obese volunteers confirmed the presence of two reversible and one nonreversible components within the skeletal muscle tissue-activity curves of the metabolism of 18F-FDG. Importantly, there was also no evidence of dephosphorylation of 18F-FDG-6-P consistent with the findings in the prior study by Bertoldo et al. (22). In addition, there are considerations other than anatomical that favor an alignment of the two reversible components occurring in series and comprised of interstitial diffusion followed by trans-membrane glucose transport. Data reported by Bergman et al. (28) indicate that interstitial diffusion of substrates and insulin play a dominant role in determining the kinetics of onset of insulin action in skeletal muscle.
In the prior study by Bertoldo et al. (17), the skeletal muscle PET model was developed using data from lean healthy volunteers studied under basal conditions and during a moderate rate of insulin infusion. In the current study, in addition to applying the model to a separate data set from lean healthy volunteers, we have extended its application in two ways. First, we extended its application for the study of obesity, which is a well known cause of skeletal muscle IR. Second, we have used the new model for data obtained at both moderate and marked insulin-stimulated conditions as well as during basal (fasting) conditions. This provided an opportunity to examine the potential interaction of glucose delivery, transport, and phosphorylation at several levels of insulin stimulation in health and IR. Thus, whether the transporters are a sole locus of control for insulin action or part of a network of distributed control (29) is central to defining the role that impaired glucose transport might have in the pathogenesis of IR.
In applying the 5K model, one clear finding was an effect of insulin on k3, the parameter pertaining to trans-membrane transport. In healthy lean volunteers, k3 increased 6- and 10-fold above fasting values. This indicates a potent effect of moderate and marked insulin concentrations to activate glucose transport in skeletal muscle. In contrast, in obese individuals studied at moderate insulin concentrations, there was a greatly diminished response or, stated otherwise, a failure of insulin activation of the rate constant for trans-membrane glucose transport. However, at marked insulin concentrations in obese individuals, the rate constant for trans-membrane glucose transport was activated to an extent not different from that found in the lean volunteers. These data clearly point to a major role for insulin-stimulated glucose transport in health and to a major role for the failure of this response in the pathogenesis of obesity-related IR.
Defects in glucose transport in skeletal muscle in obesity were previously reported by our group using dynamic PET 18F-FDG data analyzed by the classic 3K model. However, a clear insulin effect on the kinetics of glucose trans-membrane transport was not obtained using the classic model (9, 16). The K'1 parameter of the classic model describes tracer movement from plasma to tissue and includes effects of blood flow, substrate diffusion, and trans-membrane transport, but the last aspect is probably overshadowed by the other components. This is suggested by inconsistent effects of insulin to modulate values of K'1 as well as by a previously noted correlation of K'1 with blood flow (13). The K1 and k2 parameters of the new model that are ascribed to tracer delivery also did not change in response to insulin and, indeed, correlated closely with K'1 from the classic model. Thus, an exciting potential suggested for the new model is a capability to dissect out with good resolution separate parameters for 18F-FDG delivery and its trans-membrane transport.
The most striking finding in applying the classic 3K model to dynamic PET [18F]FDG data of skeletal muscle in health and IR pertained to k'3, the rate constant ascribed to the kinetics of glucose phosphorylation. Consistently, across numerous volunteers and several separate studies, insulinsensitive subjects manifested a strong stimulation of k'3, whereas insulin-resistant subjects manifested a blunted response (9, 13, 16). In the current study these findings were replicated. However, in applying the new model, the effect of insulin on the k5 parameter that is ascribed to 18F-FDG phosphorylation was substantially more subtle. Activation of k5 during moderate insulin concentrations was observed in insulin-sensitive volunteers, and this was blunted in the obese subjects with insulin resistance. However, the amplitude of change was far less dramatic than had been found in the k'3 parameter of the classic model. Our interpretation, and one that will require further investigation, is that the classic k'3 parameter as obtained from skeletal muscle probably incorporates components of insulin action on both transmembrane transport and phosphorylation. We had previously suggested this, along with the suggestion that the classic k'2 parameter reflected a failure of transport from interstitial space into muscle cells and not just failure of phosphorylation (16).
Given that the new model for compartmental analysis of dynamic PET 18F-FDG data in skeletal muscle, compared with the classic model, yields a clearer and more robust signal of trans-membrane glucose transport and suggests a substantially less robust insulin stimulation of the kinetics of glucose phosphorylation, the natural question is whether the physiological interpretations of insulin action and insulin resistance are different. Perhaps surprisingly our answer is no. The physiological interpretations of insulin action and insulin resistance that we derive from the two models are actually quite similar as an overall paradigm of how insulin influences the distribution of control within the proximal steps of glucose metabolism and how this distribution is perturbed in insulin resistance. These interpretations are based to a large extent on analysis of both the values and the changes in value of the control coefficient for transport (and its mathematical counterpoint for phosphorylation). The new model indicates that under basal conditions, the majority of control over proximal steps of glucose metabolism (70%) resides within the efficiency of glucose phosphorylation. Insulin, especially at high levels, shifts this locus in healthy, insulin-sensitive volunteers to transport, reflecting an increased efficiency in glucose phosphorylation. In the IR of obesity, the basal pattern is identical to that in lean subjects, but in response to a similar moderate level of insulin stimulation, there was a blunted redistribution of control toward transport, and in fact, in this study it would appear that control was equally distributed between transport and phosphorylation. Only at marked insulin stimulation did obese volunteers achieve a similar realignment of the distribution of control between transport and phosphorylation in the same proportions as found in lean volunteers. The blunted redistribution that occurred at moderate insulin stimulation, a value in the upper physiological range, was due to a blunted response of both the transport and phosphorylation rate constants in obesity. These are nearly the same overall interpretations concerning both insulin-sensitive and insulin-resistant responses that had been determined earlier using the classic 3K model.
In summary, this study extends the successful application of a muscle-specific model for compartmental analysis of dynamic PET 18F-FDG data across a range of insulin stimulation in lean and obese subjects. Compared with the classic model, the skeletal muscle-specific model reveals a more clearly defined effect of insulin on transmembrane glucose transport and an impairment of this response in obesity. Both the skeletal muscle-specific model and the classic model suggest an important role of glucose phosphorylation with respect to distribution of control of glucose metabolism at basal and moderate levels of insulin stimulation in lean and obese subjects.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: 2D or 3D, Two- or three-dimensional; 18F-FDG, [18F]2-deoxy-2-D-glucose; 18F-FDG-6-P, [18F]2-deoxy-2-D-glucose-phosphate; IR, insulin resistance; 3K, three-rate constant; 5K, five-rate constant; PET, positron emission tomography.
Received August 15, 2002.
Accepted December 10, 2002.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. Pendergrass, A. Bertoldo, R. Bonadonna, G. Nucci, L. Mandarino, C. Cobelli, and R. A. DeFronzo Muscle glucose transport and phosphorylation in type 2 diabetic, obese nondiabetic, and genetically predisposed individuals Am J Physiol Endocrinol Metab, January 1, 2007; 292(1): E92 - E100. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 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]
|
|
|
|
 |
|
 L. Slimani, V. Oikonen, K. Hallsten, N. Savisto, J. Knuuti, P. Nuutila, and P. Iozzo Exercise Restores Skeletal Muscle Glucose Delivery But Not Insulin-Mediated Glucose Transport and Phosphorylation in Obese Subjects J. Clin. Endocrinol. Metab., September 1, 2006; 91(9): 3394 - 3403. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Iozzo, A. Gastaldelli, M. J. Jarvisalo, J. Kiss, R. Borra, E. Buzzigoli, A. Viljanen, G. Naum, T. Viljanen, V. Oikonen, et al. 18F-FDG Assessment of Glucose Disposal and Production Rates During Fasting and Insulin Stimulation: A Validation Study J. Nucl. Med., June 1, 2006; 47(6): 1016 - 1022. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
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]
|
|
|
|
 |
|
 Y.-B. Kim, O. D. Peroni, W. G. Aschenbach, Y. Minokoshi, K. Kotani, A. Zisman, C. R. Kahn, L. J. Goodyear, and B. B. Kahn Muscle-Specific Deletion of the Glut4 Glucose Transporter Alters Multiple Regulatory Steps in Glycogen Metabolism Mol. Cell. Biol., November 1, 2005; 25(21): 9713 - 9723. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Bertoldo, J. Price, C. Mathis, S. Mason, D. Holt, C. Kelley, C. Cobelli, and D. E. Kelley Quantitative Assessment of Glucose Transport in Human Skeletal Muscle: Dynamic Positron Emission Tomography Imaging of [O-Methyl-11C]3-O-Methyl-D-Glucose J. Clin. Endocrinol. Metab., March 1, 2005; 90(3): 1752 - 1759. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. T Fueger, J. Shearer, D. P Bracy, K. A Posey, R. R. Pencek, O. P McGuinness, and D. H Wasserman Control of muscle glucose uptake: test of the rate-limiting step paradigm in conscious, unrestrained mice J. Physiol., February 1, 2005; 562(3): 925 - 935. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
