A stringent validation of mouse adipose tissue identity markers

The nature of brown adipose tissue in humans is presently debated: whether it is classical brown or of brite/beige nature. The dissimilar developmental origins and proposed distinct functions of the brown and brite/beige tissues make it essential to ascertain the identity of human depots with the perspective of recruiting and activating them for the treatment of obesity and type 2 diabetes. For identification of the tissues, a number of marker genes have been proposed, but the validity of the markers has not been well documented. We used established brown (interscapular), brite (inguinal), and white (epididymal) mouse adipose tissues and corresponding primary cell cultures as validators and examined the informative value of a series of suggested markers earlier used in the discussion considering the nature of human brown adipose tissue. Most of these markers unexpectedly turned out to be noninformative concerning tissue classification (Car4, Cited1, Ebf3, Eva1, Fbxo31, Fgf21, Lhx8, Hoxc8, and Hoxc9). Only Zic1 (brown), Cd137, Epsti1, Tbx1, Tmem26 (brite), and Tcf21 (white) proved to be informative in these three tissues. However, the expression of the brite markers was not maintained in cell culture. In a more extensive set of adipose depots, these validated markers provide new information about depot identity. Principal component analysis supported our single-gene conclusions. Furthermore, Zic1, Hoxc8, Hoxc9, and Tcf21 displayed anteroposterior expression patterns, indicating a relationship between anatomic localization and adipose tissue identity (and possibly function). Together, the observed expression patterns of these validated marker genes necessitates reconsideration of adipose depot identity in mice and humans.