The study of aquaglyceroporins expression in human subcutaneous adipose tissue will undoubtedly provide important information in understanding glycerol transport mechanisms through the adipocytes, a crucial step in the maintenance of normal adiposity. In this specific contest, we have demonstrated using different methods (RT-PCR, immunoblotting, fluorescence microscopy, and gene silencing) that AQP10 is present in human adipocytes and can be considered an alternative pathway for glycerol efflux in addition to the previously demonstrated aquaglyceroporins.
AQP7 and 10 belong to the aquaporin family, and based on their functional characteristics, to the sub-group of the aquaglyceroporins 
. Thus, their permeability was confined not only to water but also to glycerol, urea and other small solutes.
AQP7 is present in various tissues, such as gastrointestinal tract, kidney, skeletal muscle, inner ear and male reproductive system, but the adipose tissue represents the major site of AQP7 expression in humans and rodents where it reportedly promotes glycerol exit 
. Nevertheless the localization of AQP7 in adipocyte plasma membrane has been recently questioned. Nielsen’s group developed an AQP7-KO mice line to study accurately the tissue localization of AQP7 and its role in glycerol metabolism 
. Surprisingly, AQP7 protein was localized in the capillary endothelium of adipose tissue but not in adipocyte membranes.
Even if obtained in mice, the localization of AQP7 in the capillary endothelium rather than in adipocyte membranes represents a Copernican revolution in the field that opens new questions and widens the horizons.
In this direction we consider also the expression and localization of AQP7 in human adipose tissue. We found that AQP7 transcript was present not only in adipose tissue, that obviously possesses capillaries, but also in isolated and in cultured differentiated adipocytes. At protein level AQP7 expression was demonstrated by immunoblotting and using confocal immunofluorescence. Results clearly show a labeling either in the capillary endothelium or in the adipocyte membranes. Colocalization of AQP7 with CD34, well known marker of endothelial cell, confirms that AQP7 is, at least in part, associated with the plasma membrane of the small vessel of mouse adipose tissue 
. To be more convinced on AQP7 membrane localization we verified its expression and localization in cultured differentiated adipocytes. The labeling observed seems to support an AQP7 expression in adipocyte membranes in addition to the endothelial cell.
The core of this study focused however on the identification of an alternative pathway in human adipocytes plasma membrane that works together with AQP7 in glycerol outflow to adipose interstitium. The results presented here show that AQP10 is expressed in human white adipose tissue and is localized in the plasma membrane of fat cells. AQP10 mRNA was demonstrated in native subcutaneous adipose tissue, in isolated and in cultured differentiated adipocytes, even though in different amount. Immunoblotting of crude membrane confirmed also at protein level the presence of AQP10 in human adipose tissue with bands of the expected size for the monomeric and dimeric forms, as previously reported 
. Moreover, the bands were similar to those of duodenal membrane homogenates used as positive control ( left) 
. Differently from AQP7, AQP10 appears localized exclusively at the plasma membrane domain of adipocytes (). Moreover, co-localization experiments performed with anti-CD34 antibody, excluded the AQP10 presence in the plasma membrane of small vessels of adipose tissue. Interestingly, a short AQP10 variant has been recently found in the capillary endothelium of human small intestine 
. In this regard, though the expression of AQP10 has been demonstrated, until now, almost exclusively in the gastrointestinal tract, a clear-cut protein localization has not unanimously accepted. Unlike initially observed by in situ
hybridization and immunohistochemistry in villous superficial epithelial cells, a later study found two AQP10 forms expressed in granules of entero-endocrine cells and, as above stated, in the small intestinal capillary endothelium 
Functionally, aquaglyceroporins existence in human adipocyte plasma membranes has been also demonstrated by measuring osmotic water and glycerol permeability of isolated adipocytes and adipocyte plasma membranes preparations. These preparations behaved as functional osmometers sensitive to DMSO or Hg2+
treatment. Nevertheless, this result was not able to distinguish the single contribution of different aquaglyceroporins to the overall water and glycerol permeability of adipocytes since all these AQPs have been shown to be sensitive to mercurial compounds 
The expression and regulation of AQP7 was extensively studied in human adipose tissue since its primary function as a glycerol- rather then as a water-channel. Reduced glycerol outflow from the adipocytes has been suggested to be a crucial step that leads to adipocyte hypertrophy, increased adiposity until obesity onset. 
.This theory derives from exciting results obtained with AQP7-KO mice independently generated by two research groups that displayed an increase in body weight and fat accumulation in adulthood 
. These findings were confirmed also by in in vitro
results with AQP7-knockdown 3T3-L1 adipocytes 
. The results are in disagreement with those obtained using other AQP7-null mice, independently generated by Matsumura et al. 
and by Nielsen’s group 
that exhibited normal body weight, growth curves and adipose masses, normal lipolytic activity and normal glycerol plasma concentration. In humans, AQP7 gene expression was found to be down-regulated in the adipose tissue of obese vs. lean subjects 
, but unaltered in type 2 diabetes subjects 
. A common polymorphism in the promoter region of human AQP7 gene (A-953G SNP) was found to reduce AQP7 expression in adipose tissue thus increasing the risk of obesity and eventually type 2 diabetes 
. However, differently from studies on AQP7-null mice, a clear correlation between AQP7 expression and obesity failed to be observed. Indeed, human subjects homozygous for G264V mutation encoding for a non- functional AQP7 protein have normal adiposity and normal plasma glycerol concentration in basal conditions 
. This has strengthened the hypothesis of alternative glycerol channels in adipocytes 
. The results reported here can clarify the different findings observed between humans and mice: i.e., the human adipose tissue expresses AQP10 that in mice has been demonstrated to be a pseudogene 
. Results here presented demonstrate that AQP10 is regulated by lipogenic (insulin) and lipolytic (isoproterenol) stimuli thus participating to the control of fat accumulation. This regulation is similar to that described for AQP7 in 3T3-L1 adipocytes 
. These authors have also demonstrated in human subcutaneous and omental adipose tissue the presence of AQP3 and 9, with the first favouring the glycerol exit after lipolytic stimuli and the latter being constitutively present in the plasma membrane and involved in glycerol entry to lower hyperglycaemia.
In conclusion, the present study demonstrates on one hand the expression of AQP7 in human adipocyte plasma membrane and on the other hand the presence of an alternative glycerol channel, AQP10, that working with AQP3 and AQP7 ensures glycerol exit from adipocyte, thus protecting humans from obesity.
However, further investigation is required to clarify AQP10 gene expression regulation in obese and in diabetic subjects in respect to that of AQP7.