Fat and Water 1H MRI to Investigate Effects of Leptin in Obese Mice




Leptin is known to be associated with regulation of body weight and fat content. The effects of exogenous leptin on abdominal visceral (VS) and subcutaneous (SC) fat volume and hepatic fat-to-water ratio in leptin-deficient obese mice were investigated by 1H magnetic resonance imaging (MRI). Chemical shift-selected fat and water 1H MRI of control and leptin-treated mice were obtained 1 day before treatment and after 7 days of treatment (0.3 mg/kg/day). Hepatic fat-to-water ratio and VS fat volume decreased significantly with treatment, whereas SC fat volume did not change. Noninvasive measurement of fat and water content in different body regions using MRI should prove useful for evaluating new drugs for the treatment of obesity and other metabolic disorders.


Leptin, a 16 kDa protein hormone, is an important component in the regulation of body weight and fat content (1). This hormone is produced by the adipose tissue and interacts with six types of receptors, including LepRb that is present in the ventral medial nucleus of the hypothalamus. Leptin plays a key role in regulating energy metabolism and appetite via different neuronal pathways. The major circuitry for controlling appetite status begins with the lateral hypothalamus nucleus then the arcuate nucleus, that outputs signal back to the lateral hypothalamus and ventral medial lateral hypothalamus, which are the brain's feeding and the satiety centers (2). When leptin binds with the LepRb receptor, a “satiety signal” is sent to the brain.

In obese humans and rodents, leptin concentration in the plasma correlates positively with body fat content, suggesting that this kind of obesity is related to leptin insensitivity/deficiency (3,4). A mutation in the leptin gene may produce leptin insensitivity, resulting in early development of morbid obesity (>45 kg by 7–8 years of age (5,6)). An obese condition can also occur with leptin deficiency, when leptin is not produced by adipose tissue. This condition is observed in ob/ob mouse, but is not commonly found in humans. These genetically mutated mice lacks the adipose-related “satiety factor” (4).

Administration of leptin to leptin-deficient ob/ob mice results in a decrease in whole body weight in a dose-dependent manner that correlates well with an increase in the plasma leptin level within a physiological range (7,8,9). Leptin also has a great impact on metabolic activities based on the visceral (VS) or subcutaneous (SC) regions of fat distribution that contribute to the whole body weight (10,11). SC fat is found mostly underneath the skin, and can be estimated using body fat calipers. VS fat, also known as organ fat, is located inside the peritoneal cavity, packed in between the internal organs. Crude estimates of VS fat can be obtained from waist circumference or abdominal sagittal diameter. Adding the SC and VS fat gives a rough estimation of the total body adiposity (12). However, these methods provide rather unreliable measure of body fat content and are impractical in experimental animals.

1H magnetic resonance imaging (MRI) offers several methods to obtain separate fat and water images and allows quantification of fat volumes in anatomically distinct fat depots noninvasively and simultaneously (13,14,15,16,17). In addition, 1H MRI can provide estimation of fat-to-water ratio in tissues such as the liver where both are present simultaneously. The objective of this study was to evaluate the effects of exogenously administered leptin on SC and VS fat volumes, the degree of hydration of these fat reserves and hepatic fat-to-water ratio in leptin-deficient obese mice using 1H MRI.

Methods and Procedures

Animal model

The animal experiments were reviewed and approved by the Indiana University Institutional Animal Care and Use Committee. A total of 12 obese leptin-deficient mice (C57BL/∧OlaHsd-Lep; ob/ob) were subcutaneously implanted with Alzet osmotic mini-pumps containing either vehicle (5% dextrose in H2O, n = 6) or leptin (0.3 mg/kg/day, n = 6). The infusions lasted for 7 days. Fat and water MRI scans were performed 1 day before implanting the osmotic pumps and then again after 7 days of leptin or vehicle treatment. Body weight and food intake were monitored daily for both the groups. The mice were anesthetized by inhalation of 1–2% isoflurane in 100% oxygen or medical air delivered through a face mask before and during the MRI scans. The animal was positioned on a plastic cradle over a pneumatic sensor (SA Instruments, StonyBrook, NY) to monitor the respiration rate. A rectal fiber-optic probe was used to monitor the core body temperature. The animal core body temperature was maintained at 35–37 °C by blowing warm air into the magnet.

MRI protocol

All the MRI experiments were performed on a 9.4-T, 31 cm diameter horizontal bore magnet (Varian, Palo Alto, CA) that is interfaced to a Varian console. The horizontal bore system is equipped with a 12 cm diameter gradient set capable of generating 38 Gauss/cm gradient strengths in all three directions. A 38 mm diameter birdcage radio frequency coil was used for MRI data collection. Multi-slice spin-echo trans-axial images were acquired with a repetition time of 2,000 ms, echo time of 12 ms, and data matrix size of 256 × 256. Twenty-four slices of fat and water images with 2 mm slice thickness and 0.5 mm slice gap were collected with a field of view of 40 × 40 mm2. Fat and water suppressed images were acquired using the Chemical Shift Selective (CHESS) technique (18). Two 5-mm diameter reference tubes, one tube containing water and the other containing baby oil, were placed next to the animal. The reference tubes were used for ensuring adequate fat or water suppression and for signal intensity (SI) referencing (19).

Data processing and statistics

The MRI images obtained were transferred to Image Browser software (Varian application) for analysis. Three consecutive slices between the kidneys and the hind legs were selected to determine VS and SC fat volumes in the abdominal region before and after the leptin treatment. VS and SC areas were separated in the fat and water images by drawing region of interests and then applying image processing techniques such as thresholding and image subtraction for better region delineation (Figure 1). The degree of hydration of the VS and SC fat stores was evaluated from the fat and water SI relative to the SI from the baby oil and water reference tubes using the following formula (17):

Figure 1.

(a) Axial 1H fat magnetic resonance images of a control ob/ob mice showing segmentation into (b) VS and (c) subcutaneous fat images. The dotted enclosed curves in a represent ROIs for the oil reference and VS fat. ROI, region of interest; VS, visceral.


Mean ± s.e.m. fat and water SIs in the liver were determined using two rectangular region of interests (left and right side) in the liver region (Figure 5a–d). The fat-to-water SI ratios were calculated after normalizing the liver SIs with the respective SI from the reference tubes. A P value <0.05 was considered significant for comparing changes in control animals vs. leptin-treated animals and for changes before and after leptin treatment.

Figure 5.

Effect of leptin on liver fat-to-water SI ratio in ob/ob mice. 1H magnetic resonance images before (a,b) and after (c,d) leptin treatment showing changes in fat (a,c) and water (b,d) content in the liver are shown. Region of interest (white rectangles) were selected in the liver for estimating 1H fat SI and water SI. Leptin treatment decreased the fat SI and increased the water SI in the treated animals. White bars, 1 day before vehicle or leptin therapy; shaded bars, 7 days after vehicle or leptin therapy are shown in the (e) bar diagram. Leptin significantly decreased liver f at-to-water ratio compared to both the baseline and the control values. Mean ± s.e.m., n = 6. Significance: *P ≤ 0.01 (before vs. after therapy), #P ≤ 0.01 (control vs. leptin). SI, signal intensity.


Effect of leptin on body weight

Mean body weight of the ob/ob mice in control and leptin groups before treatment was 46.7 ± 0.9 g and 46.9 ± 0.8 g, respectively. In response to 5 days of leptin administration, the body weight increased to 50.2 ± 0.9 g (P < 0.05) in the control group and decreased to 43.9 ± 0.8 g (P < 0.05) in the leptin group. The final body weight significantly differed between the groups (P < 0.05).

Figure 2 shows body weight changes in control and leptin-treated groups 1, 3, and 5 days after treatment. In the control group, the body weight increased by 1.5 ± 0.2 g and 0.6 ± 0.2 g at 3 and 5 days, respectively. At the same time, the body weight of leptin group decreased (P < 0.05) by 0.3 ± 0.2 g and 1.4 ± 0.3 g, respectively. Leptin-treated animals also showed a significant decrease in their daily food intake compared to the control animals. After 5 days of treatment, daily food intake of leptin group was 2.0 ± 0.5 g compared to 5.7 ± 1.0 g for control group (P < 0.05).

Figure 2.

Effect of leptin on daily body weight changes in ob/ob mice; control—vehicle (5% dextrose in H2O) and leptin (0.3 mg/kg/day). Body weight significantly decreased on days 3 and 5 after leptin treatment. Mean ± s.e.m., n = 6. Significance: *P ≤ 0.05 (control vs. leptin).

Effect of leptin on VS and SC fat volumes

Figure 3 shows a 1H fat MRI image of a representative mouse before and after leptin treatment. Figure 4 shows the changes in the VS and SC fat volumes for control and treated animals before and after leptin treatment. In the control group, VS fat volume in the abdominal region significantly increased (2.77 ± 0.04 cc to 3.06 ± 0.11 cc; P ≤ 0.05) 7 days after vehicle treatment. In contrast, leptin significantly decreased VS fat volume 7 days after treatment (2.18 ± 0.10 cc) compared to before treatment (P < 0.05) and the control group (P < 0.01). There were no significant changes observed in the SC fat volumes for either control or leptin-treated mice. The in-plane image resolution for the MRI images was 0.156 mm. Based on the resolution, the error associated with the fat volume estimate was ∼1–1.5 %.

Figure 3.

Axial 1H fat magnetic resonance images showing the effect of leptin on VS and subcutaneous fat volume in a treated ob/ob mice (a) before and (b) after leptin treatment. The decrease in VS area is clearly visible after leptin treatment. VS, visceral.

Figure 4.

Effect of leptin on (a) VS and (b) SC fat volume in ob/ob mice. White bars, 1 day before vehicle or leptin therapy; shaded bars, 7 days after vehicle or leptin therapy. VS fat volume significantly increased over 7 days in control mice while it decreased in leptin-treated animals. SC fat volume did not change significantly in control or leptin-treated mice. Mean ± s.e.m., n = 6. Significance: *P ≤ 0.05 (before vs. after therapy), #P ≤ 0.01 (control vs. treated). SC, subcutaneous; VS, visceral.

Effect of leptin on VS and SC fat-to-water SI ratios

Before leptin treatment, the percent hydration of VS and SC regions was 13.0 ± 0.9 and 12.4 ± 0.5, respectively. The control animals did not show any change in percent hydration of VS (14.1 ± 0.4) and SC (13.6 ± 0.4) fat after the vehicle treatment. Leptin-treated animals showed a significant increase in the hydration of both VS (16.5 ± 0.5, P ≤ 0.05) and SC (15.6 ± 0.6, P ≤ 0.05) fat compared to the baseline and the control group.

Effect of leptin on hepatic fat-to-water ratio

Figure 5a–d shows 1H fat and water MRI images of the liver region of a representative mouse before and after leptin treatment. Leptin treatment decreased the fat SI from 0.171 ± 0.006 to 0.114 ± 0.006 (P < 0.01) and increased the water SI from 0.083 ± 0.002 to 0.111 ± 0.005 (P < 0.01) in the treated animals. Figure 5e shows changes in hepatic fat-to-water SI ratio in control and treated animals before and after leptin treatment. Leptin-treated animals showed a significant decrease in the hepatic fat-to-water ratio from 2.0 ± 0.1 to 1.0 ± 0.1 (P < 0.01) after 7 days of treatment. These changes in leptin-treated animals were also significantly lower (P < 0.01) compared to the control animals (1.9 ± 0.1). The changes seen in liver fat to water ratio was the most significant decrease among all the parameters measured on MRI images in the study.


Administration of exogenous leptin to leptin-deficient ob/ob mice has a profound impact on the rodents' metabolic activities that leads to a significant whole body weight loss. The MRI data presented here show that: (i) leptin plays a selective action to reduce VS fat volume than the SC fat volume in the abdominal region, (ii) leptin increases the hydration of VS and SC fat depots, and (iii) leptin produces a dramatic decrease in the fat reserves in the liver.

Computed tomography and MRI are the commonly used imaging tools to quantify VS and SC fat. Unlike computed tomography, MRI has the major advantage that it can produce images without any exposure to ionizing radiation and has demonstrated good reproducibility for measurements of the total and VS adipose tissue volumes as detected by computed tomography (20,21). VS and SC abdominal fat can be clearly distinguished from other tissues on MRI due to their relaxation and chemical shift properties. In the present study fat and water 1H MRI were acquired based on their respective chemical shift properties. A number of precautions were taken to make sure that the acquired fat and water images did not interfere with each other's signal. The reference tubes (one tube containing water and the other containing baby oil) that were placed next to the animal were used to ensure adequate fat or water suppression and for SI referencing obtained from the corresponding MRI images. As the liver contains both fat and water, a fat-to-water ratio from both images is required to quantify the fat in this compartment. Separate fat and water images also allow estimation of hydration of VS and SC fat stores.

The effects of leptin follow a number of neuronal pathways that affect various metabolic activities, leading to a decreased appetite and energy intake in ob/ob mice. One of the pathways is where the leptin binds to the LepRb present in the ventral medial nucleus of the hypothalamus, which in turn sends a satiety signal to the brain. A clue to another pathway is that when leptin is exogenously administered to leptin-deficient ob/ob mice, it inhibits the neuropeptide Y/Agouti-related protein neuron group and stimulates pro-opiomelanocortin/cocaine-andamphetamine-regulatedtranscript neurons in the arcuate nucleus. The combination of inhibition of neuropeptide Y, stimulation of pro-opiomelanocortin/cocaine-andamphetamine-regulatedtranscript neurons, and the release of corticotropin-releasing hormone, could lead to stimulating satiety and inhibit extra feeding, which could lead to a decrease in the body weight (22) as was seen in our study. A third pathway may be related leptin-induced increase in the adiponectin. This protein hormone that is secreted from the adipose tissue acts to suppress metabolic activities preventing obesity in ob/ob mice.

Leptin has a great impact on the metabolic activity based on the SC and VS distribution of fat (10). A previous study using invasive techniques has shown that administration of exogenous leptin to Sprague–Dawley rats is capable of causing a marked decrease in the VS fat as compared to SC fat (10). Selective activation of the β-3 adrenoceptor affects the VS adiposity more than the SC (23,24). The β-3 adrenoceptor's main function is to enhance lipolysis in adipose tissue. The major effect of the β-3 adrenoceptor is thought to be more selective for VS fat than SC fat and may be due, in part, to the activation of this neuronal pathway. Clinically, leptin administration to leptin-deficient patients has been shown to reduce body fat mass as well as free-fat mass as analyzed by whole body composition with dual-energy X-ray absorptiometry scanning (25). Licinio et al. demonstrated that daily SC administration of r-metHuLeptin (0.01–0.04 mg/kg; titrated according to their weight loss) produced a time-dependent decrease in body weight, fat mass, fat-free mass, and percent body fat over 18 months in three leptin-deficient patients (25). Figure 3 supports all these claims and demonstrates that leptin treatment in leptin-deficient ob/ob mice produces a significant selective reduction in VS fat whereas the SC fat remained unchanged.

The higher lipolytic activity in VS fat in comparison to SC fat is also due to the regional variation in the action of the major lipolysis regulating hormones, catecholamines and insulin. The lipolytic effect of catecholamines are more pronounced in the VS than in the SC adipose tissue (26). This is because of the increased expression and function of the β-adrenoceptor (especially β-3 adrenoceptor) and a decreased function of α-2 adrenoceptor dependent anti-lipolysis in the VS fat region. The SC fat is not as sensitive as VS fat to catecholamine-induced lipolysis due to a depletion of the β-2 adrenoceptor (27). Because the intra-abdominal VS fat has a higher fractional lipolytic rate in comparison to the SC fat region, it plays a significant role in the reduction of the whole body weight and also exerts its influence on liver metabolism (28).

The data presented show that leptin treatment can cause a significant reduction in the hepatic fat content of ob/ob mice as measured by the liver fat-to-water ratio by 1H MRI (Figure 4). There are many observations that leptin has direct effects on energy metabolism (29,30,31). In addition to its central action on food uptake, leptin also has a peripheral action in modulating fatty acid and glucose metabolism and preventing the accumulation of lipids in liver and other nonadipose tissues (32). Wein et al. demonstrated that intravenous leptin injection increased mitochondrial palmitoyl-CoA oxidation rate (+95%) and carnitine-palmitoyl transferase-1 activity (+52%) in rat liver (33). This was paralleled by significant lowering of the hepatic triglyceride content. Shimabukuro et al. showed that leptin lowered triglyceride content in pancreatic islets by preventing its formation from free fatty acids (31). The oxidation of energetic substrates, including free fatty acids, is limited in ob/ob liver mitochondria due to proton leak of mitochondrial membrane. Melia et al. found that proton leak is approximately three times greater in liver mitochondria from ob/ob mice compared to lean controls at any given membrane potential (34). Acute leptin administration restored the liver mitochondria proton leak rate of ob/ob mice to the control level leading to increase in substrate oxidation and coupling it with oxidative phosphorylation (35). Thus, our results and the observation that leptin induces a decrease in liver fat content may be explained by an activation of fatty acid oxidation and an increase in bioenergetic status.

A noninvasive evaluation of the effects of leptin on VS fat using MRI has been studied in this work. Due to the noninvasive nature of MRI, repeated fat and water MRI measurements were possible with these intact animals before and after leptin treatment. Given the significant advances in MRI techniques and the superiority of MRI over other imaging modalities in terms of spatial and temporal resolution, these techniques can be translated to clinical scanners to examine similar metabolic changes in humans and will prove useful for evaluating new drugs for obesity and other metabolic disorders.

In conclusion, leptin-deficient ob/ob mice, leptin produces a significant decrease in VS fat volume, but not in SC fat volume, an increase in the hydration of VS and SC fat depots and a large decrease in hepatic fat content. 1H MRI provides a robust method for monitoring the compartmental changes in fat and water content. Fat and water MRI should prove useful for evaluating new drugs for the treatment of obesity and other metabolic disorders.


We gratefully acknowledge Dr Jennings for his helpful discussions and editorial comments. This study was sponsored by National Institutes of Health (NIH), grant number R01EB005964.


The authors declared no conflict of interest.