Dermal white adipose tissue (DWAT) has only recently been recognized as a distinct adipocyte depot, with the potential to influence metabolism and physiology [1]. This layer of adipocytes is embedded within the dermis, and has distinct origins and physiological properties relative to other fat depots, including subcutaneous white adipose tissue (SWAT). Recent literature suggests that the DWAT layer provides natural insulation in many vertebrate species. A layer just 200 μm thick in mice is estimated to cut heat loss by two-fold [2]. Although this layer is thin, it covers the entire body surface, and therefore may comprise a significant fraction of total body fat. By reducing the activation of thermogenic programs, the thickness of DWAT could suppress brown adipose tissue activation and determine metabolic efficiency [2,3]. In this work, we describe an MRI-based approach to quantify DWAT in mice, and validate this approach using histological data.

Animals: The study included BALB/cJ and C57BL/6J wild type mice, DWAT-deficient BALB/cJ Syndecan1 (Sdc1) -/- [4] and genetically obese C57BL/6J ob/ob mice. At the time of the study, the BALB/cJ mice were 8-10 weeks old, the C57BL/6J mice were 7-11 week old and age-matched, respectively. All animals were housed at constant temperature (20-23°C) in 12 h light/dark cycles with free access to water and standard chow ad libitum. The animals were euthanized before the experiments and imaged within an hour. This work was performed in accordance with the Guide for the Care and Use of Laboratory Animals of the NIH, and approved by the local Animal Care and Use Committee.

Imaging: Imaging was performed at 3T (GE Healthcare, Waukesha, WI). Mice were placed in a prone position in an 8-channel wrist coil array (Invivo, Gainesville, FL). Fat-water separated imaging was performed using a fast spin echo (FSE) 2D multi-slice technique, including the following typical acquisition parameters: repetition time: 814ms, TE=23.8ms, 70 slices with thickness=0.6mm, spacing between slices=0.1mm, FOV=80×40mm², voxel dimensions=0.25×0.31×0.60mm³, BW=±62.5kHz, total acquisition time=7:30min, and no parallel imaging acceleration. Three echo shifts (-0.2ms, 0.6ms, and 1.4ms) were acquired to enable fat-water separation with optimal signal-to-noise ratio [5]. From these acquisitions, water-only and fat-only images were reconstructed [6].

Image Analysis: The average DWAT thickness was computed for each animal by performing region-of-interest (ROI) measurements on the fat-only images. In order to resolve partial volume effects due to DWAT thickness being possibly smaller than the in-plane spatial resolution, the DWAT thickness was calculated using the signal intensity of nearby visceral WAT as a reference. Using these ROI measurements, the average DWAT thickness in dorsal and ventral regions was estimated as follows for each animal: $$h_{DWAT} \approx \frac{s_{DWAT} \times h_{ROI}}{s_{WAT}}$$ where $$$s_{DWAT}$$$ is the average signal intensity measured in the DWAT ROI, $$$h_{ROI}$$$ is the height of the ROI covering the DWAT layer, and $$$s_{WAT}$$$ is the average signal intensity in a nearby visceral WAT depot (used as a reference to compensate for spatially-varying coil sensitivities, with the assumption that WAT is composed nearly entirely of adipose tissue).

Histological Analysis: Immediately following the imaging studies, the skin was dissected, paraformaldehyde-fixed (4%) and paraffin-embedded for evaluation.

Statistical Analysis: Data are expressed as mean +/- standard error of the mean and statistical analysis was performed with unpaired one-tailed t tests. P values less than 0.05 were considered statistically significant.

This method provided visualization of DWAT in all mice in this study. Figure 1 shows representative examples of DWAT observed both with MRI and with histology. As predicted, DWAT was thinner in Sdc1-/- mice compared to wild type, in agreement with histological analysis (Figure 2). Interestingly, DWAT was dramatically thicker in the obese mouse (Figure 3). MRI measurements agree with histology, with the following distinction. Typically, histological analyses exclude skin with anagen-stage hair follicles; during the hair cycle, follicles grow downwards into the DWAT layer, and there is a commensurate (or perhaps causative) expansion of DWAT [3,7]. To reduce the variability for any single histological section, the thickness of non-anagen skin is typically assessed. In contrast, the imaging technique collects whole-body data, and overall average thickness is readily deduced. DWAT is a layer of adipocytes embedded within the skin that has recently drawn attention for its potential to modify mammalian physiology as an independently regulated fat depot. To assess the role of DWAT in health and wellness both in humans and non-human mammals, we have developed a non-invasive fat-water imaging approach that provides an accurate assessment of the thickness of DWAT, even for the thin skin of rodent models.