Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 157 (2), 508-21

Expansion of Bone Marrow Adipose Tissue During Caloric Restriction Is Associated With Increased Circulating Glucocorticoids and Not With Hypoleptinemia

Affiliations

Expansion of Bone Marrow Adipose Tissue During Caloric Restriction Is Associated With Increased Circulating Glucocorticoids and Not With Hypoleptinemia

William P Cawthorn et al. Endocrinology.

Abstract

Bone marrow adipose tissue (MAT) accounts for up to 70% of bone marrow volume in healthy adults and increases further in clinical conditions of altered skeletal or metabolic function. Perhaps most strikingly, and in stark contrast to white adipose tissue, MAT has been found to increase during caloric restriction (CR) in humans and many other species. Hypoleptinemia may drive MAT expansion during CR but this has not been demonstrated conclusively. Indeed, MAT formation and function are poorly understood; hence, the physiological and pathological roles of MAT remain elusive. We recently revealed that MAT contributes to hyperadiponectinemia and systemic adaptations to CR. To further these observations, we have now performed CR studies in rabbits to determine whether CR affects adiponectin production by MAT. Moderate or extensive CR decreased bone mass, white adipose tissue mass, and circulating leptin but, surprisingly, did not cause hyperadiponectinemia or MAT expansion. Although this unexpected finding limited our subsequent MAT characterization, it demonstrates that during CR, bone loss can occur independently of MAT expansion; increased MAT may be required for hyperadiponectinemia; and hypoleptinemia is not sufficient for MAT expansion. We further investigated this relationship in mice. In females, CR increased MAT without decreasing circulating leptin, suggesting that hypoleptinemia is also not necessary for MAT expansion. Finally, circulating glucocorticoids increased during CR in mice but not rabbits, suggesting that glucocorticoids might drive MAT expansion during CR. These observations provide insights into the causes and consequences of CR-associated MAT expansion, knowledge with potential relevance to health and disease.

Figures

Figure 1.
Figure 1.
Circulating adiponectin does not increase during moderate CR in rabbits. Adult male rabbits were fed a control or 30% CR diet from 15 to 22 weeks of age, as described in Materials and Methods. A, Body mass was measured weekly and is presented relative to body mass at 15 weeks of age. B–G, After 7 weeks on CR or control diet, rabbits were euthanized, and fat pads, serum, WAT, and MAT were isolated. B, WAT masses were recorded at necropsy. C, Serum leptin concentrations, as determined by ELISA. D and E, Total RNA was isolated from iWAT (D) and tibial MAT (E). Expression of the indicated transcripts was determined by qPCR and normalized to Ppia expression. F, Immunoblot of total adiponectin in sera from 22-week-old rabbits. E, Densitometry was used to quantify serum adiponectin from F. Data in A are reported as mean ± SD of 5 control and 6 CR rabbits. All other graphs are box and whisker plots. Statistically significant differences between control and CR rabbits are indicated as follows: *, P < .05; **, P < .01; ***, P < .001.
Figure 2.
Figure 2.
BM adiposity does not increase during moderate CR in rabbits. Control and moderate CR rabbits (described in Figure 1) were euthanized, and humeri, radii, ulnae, tibiae, and femurs were removed. A and B, Representative images of bisected humeri (A) or femurs (B); scale bar, 1 cm. C–E, Whole, intact BM was isolated from one femur of each rabbit, followed by isolation of total RNA (C), protein (D), or lipid (E). In C, expression of the indicated transcripts was determined by qPCR and normalized to Ppia expression. In D, perilipin A expression was determined by immunoblotting, with α-tubulin used as a loading control. In E, total triacylglycerol was isolated by TLC and the concentration determined by an enzymatic assay. F, Schematic showing the sites from which each sample of RM or MAT was isolated. G, Adipocyte size distribution in the indicated RM, MAT, or WAT samples was determined by quantitative histomorphometry; median adipocyte size was then determined. Data in C and E represent 5 control and 6 CR rabbits and are shown as box and whisker plots. In C, statistically significant differences between control and CR rabbits are indicated by * (P < .05). Data in G are reported as mean ± SD of the next numbers of samples: femur RM: 5 control, 6 CR; humerus RM: 5 control, 6 CR; tibia RM: 3 control, 5 CR; femur MAT: 5 control, 5 CR; humerus MAT: 4 control, 6 CR; tibia MAT: 5 control, 6 CR; radius MAT: 5 control, 6 CR; ulna MAT: 4 control, 6 CR; gWAT, iWAT, or pWAT: 5 control and 6 CR. In G, statistically significant differences in median adipocyte size were assessed by two-way ANOVA. Lowercase letters indicate statistical significance for the control tissues, whereas uppercase letters are used for the CR tissues; samples that do not share a common letter are significantly different from each other (P < .05). Significant effects of CR, within each tissue type, are indicated in Supplemental Figure 2.
Figure 3.
Figure 3.
Circulating adiponectin does not increase during extensive CR in rabbits. Male rabbits were fed ad libitum (control) or at 40% of ad libitum food intake (CR) from 6 to 13 weeks of age. A, Body mass was measured weekly. B–F, After 7 weeks of CR or control diet, rabbits were euthanized and fat pads, serum, and bones were isolated. B, WAT masses were recorded at necropsy. C, Serum leptin concentrations, as determined by ELISA. D, Lengths of the indicated bones were recorded at necropsy. E, Immunoblot of total adiponectin in sera from 13-week-old rabbits. F, Densitometry was used to quantify serum adiponectin from E. Data in A are reported as mean ± SD of 6 control and 6 CR rabbits. All other graphs are box and whisker plots. Statistically significant differences between control and CR rabbits are indicated as described for Figure 1.
Figure 4.
Figure 4.
BM adipocyte size is decreased during extensive CR in rabbits. Control and extensive CR rabbits (described in Figure 3) were euthanized, and humeri, radii, ulnae, tibiae, and femurs were removed. A and B, Representative images of bisected humeri (A) or tibiae (B); scale bar, 1 cm. C, Median adipocyte size in the indicated RM, MAT, or WAT samples was determined by quantitative histomorphometry, as described for Figure 2G. Because of the extent of MAT and WAT loss, from some CR rabbits we were unable to detect any MAT for further analysis. Thus, data in C are reported as mean ± SD of the next numbers of samples: femur RM: 6 control, 4 CR; humerus RM: 5 control, 4 CR; tibia RM: 5 control, 4 CR; femur MAT: 4 control, 2 CR; humerus MAT: 4 control, 4 CR; tibia MAT: 6 control, 4 CR; radius MAT: 5 control, 5 CR; ulna MAT: 6 control, 5 CR; gWAT: 5 control, 3 CR; iWAT: 6 control, 1 CR; pWAT: 5 control, 1 CR. Because femur MAT for the CR group is from only 2 rabbits, the SD of this group represents 0.7071 times the range of the 2 data points. Significant differences are indicated as described for Figure 2G. Data for iWAT and pWAT are from only one CR rabbit, and data for femur MAT are from only 2 CR rabbits; hence, ANOVA could not be used to assess statistical significance for these samples owing to uncertainty over the normality of data distribution.
Figure 5.
Figure 5.
In female mice CR increases MAT without decreasing circulating leptin. Male and female C57BL/6J mice were fed ad libitum or a 30% CR diet from 9 to 15 weeks of age. A and B, Tibiae from 15-week-old mice were stained with osmium tetroxide followed by μCT analysis. A, Representative μCT scans of osmium tetroxide-stained tibiae. MAT appears as darker regions within each bone. B, MAT volume within each tibial region was determined from μCT scans. C and F, Body composition of 15-week-old live mice was determined by NMR. D and G, Masses of the indicated tissues were recorded at necropsy. E and H, Blood was sampled from the lateral tail vein of 15-week-old live mice. Serum was isolated, and leptin concentrations were determined by ELISA. Data in C and D and F and G are reported as mean ± SD of the next numbers of mice: male control, n = 6; male CR, n = 7; female control, n = 6; female CR, n = 5. All other graphs are box and whisker plots. For each sex, statistically significant differences between control and CR animals are reported as described for Figure 1.
Figure 6.
Figure 6.
MAT expansion during CR is associated with changes in circulating glucocorticoids. C57BL/6J mice (A and B) or New Zealand White rabbits (C–F) were fed control or CR diets, as described in Figures 1–5. Blood was sampled at the end of the CR protocols, and concentrations of total corticosterone and cortisol were determined by ELISA. Data are presented as box and whisker plots. Within each group (male mice; female mice; moderate CR rabbits; extensive CR rabbits) statistically significant differences between control and CR animals are reported as described for Figure 1.

Similar articles

See all similar articles

Cited by 31 PubMed Central articles

See all "Cited by" articles

Publication types

Feedback