Introduction

Obesity is thought to be a causal pathway to insulin resistance and thus type 2 diabetes; however, not all people with obesity become insulin resistant¹. Indeed, the causal mechanisms behind the pathogenesis of insulin resistance and type 2 diabetes are unknown. Interrogation of the dichotomy of those with obesity who develop insulin resistance and those who do not might inform on the unknown pathogenesis of insulin resistance and type 2 diabetes. Recent studies suggest that adipose hypoxia is a primary pathway to systematic insulin resistance and type 2 diabetes^2-3– a mechanism which might account for the divergence between “healthy” and insulin-resistant people with obesity. The adipose-hypoxia-driven insulin resistance hypothesis has typically been interrogated postmortem by histologic observation and assay of hypoxia biomarkers (e.g., HIF-1α and pimonidazole). Additionally, invasive platinum Clark-type electrodes and fiber optic probes can be surgically implanted into tissue to provide a direct measure of tissue oxygenation, though these measurements are contaminated by the non-homeostatic response to the local trauma induced by the implantation of the device. A non-invasive/non-destructive method for measuring adipose tissue pO₂ would be a significant advance in the exploration of the hypoxia-driven insulin resistance hypothesis. To this end, the paramagnetic nature of molecular oxygen (O₂) can, in principle, be exploited to directly quantify the O₂ content of adipose tissue via simply measuring the longitudinal relaxation rate constant, R₁, of lipid-associated resonances in adipose tissue. The motivation of the current work is to employ a porcine lard phantom platform as an adipose tissue surrogate for the development of a direct, non-invasive and quantitative measure of adipose tissue oxygenation and its primary confound, temperature, using MR. Such a technique which would provide critical insight into the pathogenesis of type 2 diabetes.

Methods

Porcine lard phantom preparation: Phantoms were prepared in 8-mm NMR tubes by melting pure porcine lard at 45°C that was bubbled with N₂/O₂ gas mixtures to achieve different oxygen tension (pO₂) levels. The samples were cooled to desired temperatures (34, 37 or 40°C) prior to MR scans and pO₂ was monitored via LICOX^® device. Proof-of-principle in-vivo MR: IR-PRESS (see below) data were collected in the white adipose of an anesthetized healthy C57Bl/6 mouse under temperature control (37-40°C) at discrete adipose pO₂ levels (as determined by OxyLite probe). Adipose was modulated by varying breathing gas O₂ content (balance N₂) from 12.5% O₂ to 100% O₂. MR experiments: To obtain data for each pO₂ and T, an in-house-developed Inversion Recovery Point RESolved Spectroscopy (IR-PRESS) sequence was employed at 4.7T: TR/TE=4500/13 ms, 64 exponential-spaced inversion times from 10 to 2000 ms, NS=1, DS=1. Inversion recovery data were analyzed and modeled as a three parameter mono-exponential recovery to a constant, using in-house-developed Bayesian probability theory-based software⁴. Model: In this work we will experimentally change pO₂ and temperature (T) of porcine lard phantoms in order to quantify the relationship between pO₂ and R₁ (r_1,pO2) and T and (r_1,T), respectively. Once resolved, these coefficients can be used to convert the R₁ of two lipid-associated resonances into non-invasive measures of pO₂ and T according to Eq. 1: $$ R_{1,obs}=R_{1,ref,n}+r_{1,pO_{2},n}\cdot pO_{2}-r_{1,T,n}\cdot T $$ where R_1,obs is the measured longitudinal relaxation rate constant of a given resonance, R_1,ref is a constant at pO₂=0 mmHg and T=37°C.

Results

Porcine lard is an ideal in-vitro surrogate of adipose tissue; both porcine lard and white adipose of mice have similar chemical composition of triglycerides resulting in nearly identical NMR spectra (Figure 1). The expected linear relationships between R_1,obs and pO₂ and T are quantified (see Table 1). Further, the linear relationship between pO₂ and R₁ is characterized in-vivo (Figure 3) and the r_1,pO2 and R_1,ref are calculated to be 1.9x10^-3 mmHg^-1sec^-1 and 2.11 sec^-1, respectively, and are very close to those values obtained from the methylene peak of lard samples (peak 2, Table 1), indicating that lard is an ideal in-vitro candidate of adipose tissue.

Discussion and Conclusion

A non-invasive and quantitative measure of adipose pO₂ and its principle confound, temperature, is described. As anticipated (Eq. 1), we observed a positive linear relationship between the R_1,obs and pO₂ (R²>0.85, Figure 2A) and a negative linear relationship between R_1,obs and temperature (R²>0.99, Figure 2B). A proof-of-principle in-vivo experiment on subcutaneous white adipose of a mouse shows very similar spectral features between white adipose and lard and a very similar r_1,pO2 for the methylene peak, suggest that the principles learned from the lard phantom experiments will translate in-vivo. The phantom studies described herein will inform ongoing pre-clinical studies of metabolic disease pathways.

Acknowledgements

We would like to thank our colleagues Drs. Joseph Ackerman and Joel Garbow for their invaluable scientific and professional guidance

References

1. Pajunen P, et al. Metabolically healthy and unhealthy obesity phenotypes in the general population: the FIN-D2D Survey. BMC Public Health. 2011;11:754. 2. Trayhurn P. Hypoxia and adipose tissue function and dysfunction in obesity. Physiol Rev. 2013;93(1):1-21. 3. Ye J. Emerging role of adipose tissue hypoxia in obesity and insulin resistance. Int J Obes. 2009;33(1): 54–66. 4. Marutyan KR and Bretthorst GL. The Bayesian Analysis Software developed at Washington University. Bayesian inference and maximum entropy methods in science and engineering. American Institute of Physics, Melville, NY, 2009, pp. 368-381.