The absence of transcriptional regulation for adiponectin by insulin deprivation is also coherent with the unchanged concentration of the secreted protein in the plasma of diabetic rats compared to controls. The extracellular matrix of adipose tissue may play an important role in regulating changes in cell size. Previous studies have shown that adipose collagen VI transcripts are PF-4217903 increased in obese subjects and fibronectin is known to modulate adipose cell shape and to interfere with the regulation of lipolysis. After insulin deprivation, we found a reduced transcription level of both genes which can be considered as a direct effect of insulin privation on protein synthesis, thus may favor reduction in adipose cell size. The most astonishing observation is that cell size distribution returned close to the control values in response to insulin treatment. An outstanding observation is the increased proportion of small adipose cells in scWAT after insulin treatment corresponding to a significant reduction in adipose cell size. Estimation of adipose cell number per depot shows that cell number decreased only in scWAT in response to insulin deprivation. Counting of cells by Coulter counter may not identify cells below a certain detection limit and such cells may not be detected and thus do not show up on the screen. Our results mean that adipose cell numbers have decreased but it does not necessarily mean that also the total cell number decreased. It only means that adipose cells reached size values lower than the detection limit but are most likely still present in adipose depot, as suggested. Our results suggest that most of the changes observed in adipose tissue result from changes in adipose cell size. This is fully coherent with recent studies using doxycycline-inducible mature adipocyte–specific tracing system in mouse, since the authors found that it is mainly adipose cell hypertrophy that explains increased adipose tissue mass in the short-term response to high fat diet. Furthermore, adipocytes are fully differentiated cells that are incapable of mitotic division. This suggests that adipocyte hyperplasia arises as a result of differentiation of stem cells. Although stem cell proliferation may likely occur in response to insulin replacement, this seems very limited since adipose cell number did not significantly increased in the different WAT depots. Although the site of insulin injection may have influenced the effects on scWAT, we do not think that the effects of insulin on scWAT are overestimated, indeed, among the 17 genes which transcripts were quantified, insulin increased the transcription of 6 of them in eWAT and inhibited the transcription for 3, 2 and 6 genes, respectively. Overall, the number of genes whose transcription was altered in absence of insulin is very similar in the different adipose tissues. Whether insulin works directly at the transcriptional level or indirectly through the intermediate metabolism is an open question. SREBP1c was reported to mediate partly the transcriptional effects of insulin. Indeed, in vitro, SREBP1c enhances the transcriptional activity of PPARc, thus increasing the proportion of cells undergoing adipose differentiation.