Seminar on Bone Marrow Fat by Mark C. Horowitz

Seminar on Bone Marrow Fat by Mark C. Horowitz  -

On 4 December 2017, Professor Moustapha Kassem, Head of the KMEB research unit, Odense University Hospital, hereby invites all interested to attend the seminar entitled "Bone Marrow Fat: the 4th Type of Fat" by Professor Mark Charles Horowitz, Yale University School of Medicine. 


10:00-11:00 AM

Room 111, J.B. Winsløws Vej 25
DK-5000 Odense C


Registration is not required - all interested are welcome to attend.


Marrow adipose tissue (MAT) was identified in the bone marrow (BM) more than a century ago and until very recently was thought to be nothing more than filler. Although recent data suggests that MAT has characteristics of brown adipose tissue (BAT) or beige adipocytes, little is known about the origin, development, and function of MAT because the BM adipocyte precursor/progenitor cell remains unknown.

We have performed lineage tracing of BM adipocytes following induction of BM adipogenesis with either rosiglitazone-enriched (rosi) diet or irradiation of double fluorescent mT/mG reporter mice (which harbor a floxed membrane targeted tdTomato cassette (mT) upstream of a Cre-responsive membrane targeted eGFP cassette (mG)) to determine the ontogeny of BM adipocytes and identity of BM adipocyte progenitor cells. Following either irradiation or rosi diet, all BM adipocytes were eGFP+ and therefore traced by expression of Cre-recombinase in Adiponectin-Cre:mT/mG and Osx1-Cre:mT/mG mice. In contrast, adipocytes were not traced (tdTomato+) in Vav1-Cre:mT/mG or Myf5-Cre:mT/mG mice.

B6 mice fed a methionine restricted (MR) diet show a pronounced induction of multilocular beige adipocytes in their inguinal depot (IWAT) and a concomitant increase in MAT as compared to mice on control diet. However, unlike the IWAT, none of the marrow adipocytes were multilocular. Using immunohistochemistry, IWAT from the MR mice stained with anti-UCP-1 confirmed the presence of beige adipocytes. In contrast, MAT from those same mice was uniformly UCP-negative.

We previously demonstrated that all adipocytes in classical white adipose tissue (WAT) depots were tracedin PdgfRα-Cre:mT/mG mice due to expression of PdgfRα in WAT precursor cells. In contrast, between 50-70% of BM adipocytes were eGFP+ in the same mice. To better understand the relationship between MAT and WAT we examined PdgfRα-cre:insulinRfl/fl mice. These mice were found to have a marked increase in MAT and a paucity of WAT, suggesting that the requirement of insulin signaling for adipocyte formation is another distinguisher between MAT and WAT.

Mouse tibias and femora can be made deficient in MAT from birth using a genetic approach. These bones have increased bone mineral density and are biomechanically stronger.

Together, the uniform tracing of BM adipocytes by the endosteal cell/osteoblast promoter Osx1, the lack of tracing of BM adipocytes by Myf5 (which traces all brown adipocytes), the absence of multilocular, UCP-1 positive marrow adipocytes, and the formation of MAT in the absence of insulin receptor indicate that the MAT depot is distinct from BAT, WAT, or beige fat. MAT is not “just filler” but has demonstrable functions.


For further information or questions please contact Weimin Qiu, KMEB,