Limits...
Human adipose cells in vitro are either refractory or responsive to insulin, reflecting host metabolic state.

Lizunov VA, Stenkula KG, Blank PS, Troy A, Lee JP, Skarulis MC, Cushman SW, Zimmerberg J - PLoS ONE (2015)

Bottom Line: Two statistically-defined populations best describe the observed cellular heterogeneity, representing the fractions of refractive and responsive adipose cells.Thus, a two-component model best describes the relationship between cellular refractory fraction and subject SI.Since isolated cells exhibit these different response characteristics in the presence of constant culture conditions and milieu, we suggest that a physiological switching mechanism at the adipose cellular level ultimately drives systemic SI.

View Article: PubMed Central - PubMed

Affiliation: Program in Physical Biology, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, United States of America.

ABSTRACT
While intercellular communication processes are frequently characterized by switch-like transitions, the endocrine system, including the adipose tissue response to insulin, has been characterized by graded responses. Yet here individual cells from adipose tissue biopsies are best described by a switch-like transition between the basal and insulin-stimulated states for the trafficking of the glucose transporter GLUT4. Two statistically-defined populations best describe the observed cellular heterogeneity, representing the fractions of refractive and responsive adipose cells. Furthermore, subjects exhibiting high systemic insulin sensitivity indices (SI) have high fractions of responsive adipose cells in vitro, while subjects exhibiting decreasing SI have increasing fractions of refractory cells in vitro. Thus, a two-component model best describes the relationship between cellular refractory fraction and subject SI. Since isolated cells exhibit these different response characteristics in the presence of constant culture conditions and milieu, we suggest that a physiological switching mechanism at the adipose cellular level ultimately drives systemic SI.

Show MeSH

Related in: MedlinePlus

Representative “Bee swarm” plots (a) and empirical cumulative distribution functions (b) for basal (black) and insulin-treated (red) GSV mobility rates measured in adipose cells isolated from subjects with varying degrees of insulin sensitivity reveals two populations at the individual subject level.Dashed lines represent cluster boundaries calculated using the K-means clustering function. At low SI both the basal and insulin stimulated distributions are fit by a single, basal function. With increasing SI, increasing number of cells shift from the basal into the insulin-stimulated state characterized by low GSV mobility (less than 4 Trajectories/100 μm2/min).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4359092&req=5

pone.0119291.g005: Representative “Bee swarm” plots (a) and empirical cumulative distribution functions (b) for basal (black) and insulin-treated (red) GSV mobility rates measured in adipose cells isolated from subjects with varying degrees of insulin sensitivity reveals two populations at the individual subject level.Dashed lines represent cluster boundaries calculated using the K-means clustering function. At low SI both the basal and insulin stimulated distributions are fit by a single, basal function. With increasing SI, increasing number of cells shift from the basal into the insulin-stimulated state characterized by low GSV mobility (less than 4 Trajectories/100 μm2/min).

Mentions: To perform a quantitative analysis of the basal and insulin-stimulated distributions in GSV mobility and fusion rates, the pooled adipose cell data were represented as empirical cumulative distribution functions (CDF, Fig. 4 C, D) and fit to model CDFs with one or more Gaussian distributions. The basal distributions were characterized by zero-truncated Gaussian cumulative distributions (Fig. 4 C,D, black dotted lines), while the insulin-stimulated distributions were best characterized by two zero-truncated Gaussian cumulative distributions, one of which matched the basal distribution parameters (Fig. 4 C,D red dotted lines). If our hypothesis is correct, we would predict that the adipose cell data from individual subjects should be described by one (basal) or two (basal plus insulin-stimulated) Gaussian cumulative distributions. However, the distribution models comprise too many parameters to accurately describe the individual data sets. To resolve this problem, the means and standard deviations from the pooled data fits were used as fixed parameters, reducing our model to only one free parameter, the fraction of adipose cells, in each data set, present in the basal state. We next fit the pooled cellular data from all 19 subjects to determine the fraction of refractory adipose cells and verified if these models satisfactorily describe the data over the entire SI range. Scatter plots and empirical CDFs, and their model fits are presented in Figs. 5 and 6 for three representative subjects with low, medium, and high SI (for Mobility (Fig. 5), fractions (95% confidence) = 1 (.15), 0.24 (.13), 0 (.22), R^2 = 0.91, 0.87, and 0.88, respectively and for Fusion (Fig. 6), fractions (95% confidence) = 1 (.21), 0.38 (.21), 0.21 (.13), R2 = 0.82, 0.85, and 0.86, respectively). With increasing SI, increasing insulin-stimulated distribution fractions are observed; however, consistent with previous observations, even in the most insulin-sensitive subjects some adipose cells are observed to exhibit basal-like GSV mobility and fusion rates (Fig. 1).


Human adipose cells in vitro are either refractory or responsive to insulin, reflecting host metabolic state.

Lizunov VA, Stenkula KG, Blank PS, Troy A, Lee JP, Skarulis MC, Cushman SW, Zimmerberg J - PLoS ONE (2015)

Representative “Bee swarm” plots (a) and empirical cumulative distribution functions (b) for basal (black) and insulin-treated (red) GSV mobility rates measured in adipose cells isolated from subjects with varying degrees of insulin sensitivity reveals two populations at the individual subject level.Dashed lines represent cluster boundaries calculated using the K-means clustering function. At low SI both the basal and insulin stimulated distributions are fit by a single, basal function. With increasing SI, increasing number of cells shift from the basal into the insulin-stimulated state characterized by low GSV mobility (less than 4 Trajectories/100 μm2/min).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4359092&req=5

pone.0119291.g005: Representative “Bee swarm” plots (a) and empirical cumulative distribution functions (b) for basal (black) and insulin-treated (red) GSV mobility rates measured in adipose cells isolated from subjects with varying degrees of insulin sensitivity reveals two populations at the individual subject level.Dashed lines represent cluster boundaries calculated using the K-means clustering function. At low SI both the basal and insulin stimulated distributions are fit by a single, basal function. With increasing SI, increasing number of cells shift from the basal into the insulin-stimulated state characterized by low GSV mobility (less than 4 Trajectories/100 μm2/min).
Mentions: To perform a quantitative analysis of the basal and insulin-stimulated distributions in GSV mobility and fusion rates, the pooled adipose cell data were represented as empirical cumulative distribution functions (CDF, Fig. 4 C, D) and fit to model CDFs with one or more Gaussian distributions. The basal distributions were characterized by zero-truncated Gaussian cumulative distributions (Fig. 4 C,D, black dotted lines), while the insulin-stimulated distributions were best characterized by two zero-truncated Gaussian cumulative distributions, one of which matched the basal distribution parameters (Fig. 4 C,D red dotted lines). If our hypothesis is correct, we would predict that the adipose cell data from individual subjects should be described by one (basal) or two (basal plus insulin-stimulated) Gaussian cumulative distributions. However, the distribution models comprise too many parameters to accurately describe the individual data sets. To resolve this problem, the means and standard deviations from the pooled data fits were used as fixed parameters, reducing our model to only one free parameter, the fraction of adipose cells, in each data set, present in the basal state. We next fit the pooled cellular data from all 19 subjects to determine the fraction of refractory adipose cells and verified if these models satisfactorily describe the data over the entire SI range. Scatter plots and empirical CDFs, and their model fits are presented in Figs. 5 and 6 for three representative subjects with low, medium, and high SI (for Mobility (Fig. 5), fractions (95% confidence) = 1 (.15), 0.24 (.13), 0 (.22), R^2 = 0.91, 0.87, and 0.88, respectively and for Fusion (Fig. 6), fractions (95% confidence) = 1 (.21), 0.38 (.21), 0.21 (.13), R2 = 0.82, 0.85, and 0.86, respectively). With increasing SI, increasing insulin-stimulated distribution fractions are observed; however, consistent with previous observations, even in the most insulin-sensitive subjects some adipose cells are observed to exhibit basal-like GSV mobility and fusion rates (Fig. 1).

Bottom Line: Two statistically-defined populations best describe the observed cellular heterogeneity, representing the fractions of refractive and responsive adipose cells.Thus, a two-component model best describes the relationship between cellular refractory fraction and subject SI.Since isolated cells exhibit these different response characteristics in the presence of constant culture conditions and milieu, we suggest that a physiological switching mechanism at the adipose cellular level ultimately drives systemic SI.

View Article: PubMed Central - PubMed

Affiliation: Program in Physical Biology, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, United States of America.

ABSTRACT
While intercellular communication processes are frequently characterized by switch-like transitions, the endocrine system, including the adipose tissue response to insulin, has been characterized by graded responses. Yet here individual cells from adipose tissue biopsies are best described by a switch-like transition between the basal and insulin-stimulated states for the trafficking of the glucose transporter GLUT4. Two statistically-defined populations best describe the observed cellular heterogeneity, representing the fractions of refractive and responsive adipose cells. Furthermore, subjects exhibiting high systemic insulin sensitivity indices (SI) have high fractions of responsive adipose cells in vitro, while subjects exhibiting decreasing SI have increasing fractions of refractory cells in vitro. Thus, a two-component model best describes the relationship between cellular refractory fraction and subject SI. Since isolated cells exhibit these different response characteristics in the presence of constant culture conditions and milieu, we suggest that a physiological switching mechanism at the adipose cellular level ultimately drives systemic SI.

Show MeSH
Related in: MedlinePlus