Impaired Tissue Oxygenation in Metabolic Syndrome Requires Increased Microvascular Perfusion Heterogeneity

Abstract

Metabolic syndrome (MS) in obese Zucker rats (OZR) is associated with impaired skeletal muscle performance and blunted hyperemia. Studies suggest that reduced O₂ diffusion capacity is required to explain compromised muscle performance and that heterogeneous microvascular perfusion distribution is critical. We modeled tissue oxygenation during muscle contraction in control and OZR skeletal muscle using physiologically realistic relationships. Using a network model of Krogh cylinders with increasing perfusion asymmetry and increased plasma skimming, we predict increased perfusion heterogeneity and decreased muscle oxygenation in OZR, with partial recovery following therapy. Notably, increasing O₂ delivery had less impact on VO₂ than equivalent decreases in O₂ delivery, providing a mechanism for previous empirical work associating perfusion heterogeneity and impaired O₂ extraction. We demonstrate that increased skeletal muscle perfusion asymmetry is a defining characteristic of MS and must be considered to effectively model and understand blood-tissue O₂ exchange in this model of human disease.

Keywords: Blood flow control; Microcirculation; Oxygenation; Rodent models of obesity; Simulation.

Fig. 1

Previous studies establish impaired oxygen diffusion and microvascular perfusion heterogeneity in the OZR. a Femoral artery blood flow is significantly reduced in the OZR during 5 Hz contractions. This defect is not affected by treatment with TEMPOL. b Oxygen extraction is significantly reduced in the OZR during 5 Hz contractions. This defect is reversed by treatment with TEMPOL. c Oxygen diffusion capacity, a determinant of oxygen extraction, is a determinant of muscle fatigue for both LZR and OZR across multiple pharmacological treatments. d Schematic illustrating the parameter γ used to quantify microvascular perfusion heterogeneity. e Microvascular perfusion heterogeneity in 4a and 5a arterials is significantly increased in the OZR. This defect is reversed by treatment with TEMPOL. a–c Recreated with permissions from Frisbee et al. Exp Physiol 2011 [11]. d, e Recreated with permissions from Frisbee et al. Am J Physiol 2011 [17]

Fig. 2

Visualizations of simulations performed. a A schematic illustration of the idealized microvascular network used in the present study. Note that 4 vessel generations are visualized, whereas 15 vessel generations were used in the simulation. Please see text for details. b A schematic illustration of the Krogh cylinder-type model used for simulation of single-capillary oxygen transport in this study

Fig. 3

Heatmaps visualizing lookup tables (LUTs) of venous oxygenation as a function of capillary hematocrit and relative blood flow. a LUT of simulated venous oxygenation in the LZR at rest. b LUT of simulated venous oxygenation in the LZR during 5 Hz contractions. Note that venous oxygenation is substantially decreased (i.e., oxygen extraction is increased) relative to rest given similar flow and hematocrit. c LUT of simulated venous oxygenation in the OZR at rest. d LUT of simulated venous oxygenation in the OZR during 5 Hz contractions. Note that venous oxygenation is substantially less than at rest and marginally less than in the OZR given similar flow and hematocrit

Fig. 4

Heatmaps visualizing lookup tables (LUTs) of skeletal muscle oxygenation as a function of capillary hematocrit and relative blood flow. a LUT of simulated muscle oxygenation in the LZR at rest. b LUT of simulated muscle oxygenation in the LZR during 5 Hz contractions. Note that muscle oxygenation is substantially decreased relative to rest given similar flow and hematocrit. c LUT of simulated muscle oxygenation in the OZR at rest. d LUT of simulated muscle oxygenation in the OZR during 5 Hz contractions. Note that muscle oxygenation is substantially less than at rest and marginally less than in the OZR given similar flow and hematocrit

Fig. 5

Sensitivity analysis reveals that the effects of varying vessel diameters are negligible relative to the effects of varying flow distributions. a Variance in simulated CvO₂ increases quadratically with increasing variance of vessel diameters. At the level of diameter variance (10%) required to explain the maximum degree of perfusion heterogeneity considered in this manuscript (γ = 0.7), the effects of diameter variance without flow heterogeneity are still several orders of magnitude less than the effects of flow heterogeneity without diameter variance. b Variance in simulated PmO₂ increases quadratically with increasing variance of vessel diameters. This effect is several orders of magnitude smaller than the effects of flow heterogeneity. c Variance in simulated VO₂ increases quadratically with increasing variance of vessel diameters. This effect is several orders of magnitude smaller than the effects of flow heterogeneity

Fig. 6

Differences in venous oxygenation, muscle oxygenation, and perfusion heterogeneity can be predicted using a simulation of microvascular blood flow and oxygen transport. a Venous oxygenation increases with increasing perfusion heterogeneity in both the LZR and the OZR. The intersection between simulated venous oxygenation (light, dashed lines) and empirically measured venous oxygenation (bold lines) can be used to predict the degree of perfusion heterogeneity. Our model correctly predicts greater perfusion heterogeneity in the OZR and partial correction of perfusion heterogeneity with TEMPOL treatment. b The intersection between simulated muscle oxygenation (light, dashed lines) and model-predicted perfusion heterogeneity (bold lines) can be used to predict muscle oxygenation. Our model correctly predicts reduced muscle oxygenation in the OZR and also predicts partial correction of muscle oxygenation with TEMPOL treatment. c The intersections of simulated oxygen uptake (light, dashed lines) with empirically measured oxygen uptake (bold lines) and with model-predicted oxygen uptake without perfusion heterogeneity in the OZR (light, dotted line) can be used to predict the relative contributions of various oxygen transport parameters to the observed oxygen uptake defect. Our model predicts that perfusion heterogeneity in small (3a-5a) arterioles plays a major role in reduced oxygen uptake in the OZR

Fig. 7

Analysis of perfusion heterogeneity effects at a single bifurcation reveals the mechanisms by which perfusion heterogeneity interferes with oxygen transport. a Increasing flow through one capillary while subtracting the same amount of flow from another capillary results in a greater decrease in oxygen uptake by the under-perfused capillary than the corresponding increase in oxygen uptake by the over-perfused capillary. This effect occurs independently of hematocrit effects. b Increasing hematocrit in one capillary while subtracting the same number of RBCs from another capillary results in a greater decrease in oxygen uptake by the low-hematocrit capillary than the corresponding increase in oxygen uptake by the high-hematocrit capillary. This effect occurs independently of flow effects. c Capillary flow and hematocrit are correlated at the single-capillary level, and flow heterogeneity results in hematocrit heterogeneity. These consequences of the plasma-skimming effect cause the effects of flow heterogeneity and hematocrit heterogeneity to synergize in reducing oxygen uptake under conditions of microvascular perfusion heterogeneity