Obesity is an enlargement of adipose tissue to store excess energy intake. Hyperplasia 1(cell number increase) and hypertrophy (cell size increase) are two possible growth mechanisms. Adipose tissue obesity phenotypes are influenced by diet as well as genetics, or by their interaction. There is an extensive literature on adipose tissue growth in normal and abnormal development, characterizing the state of the tissue in terms of the mean cell size and cell number. Hyperplastic growth appears only at early stages in adipose tissue development. Hypertrophy occurs prior to hyperplasia to meet the need for additional fat storage capacity in the progression of obesity. However, it has proven difficult to understand how diet and genetics specifically affects hyperplasia and/or hypertrophy of adipose cells, because of limited data about adipose tissue growth.? Beyond these studies, it has recently become possible to measure cell-size distributions precisely. This detailed information, compared with the mean cell size and total cell number, can be used to compute many size-related quantities that permit a finer characterization of the adipose tissue growth process. Cumulants of the cell-size distribution can be used to compute other physiological quantities such as the volume-weighted mean cell size. The cell-size distribution can be used to estimate total cell number within a fat pad from its mass. Furthermore, it is believed that some specific metabolic properties, e.g., insulin resistance and adipokine secretion, depend on the precise cell-size distribution rather than the mean cell size. Indeed, several studies have addressed the change of the size distribution of adipose cells under various conditions in chick embryo development, lean and obese Zucker rats, partially lipectomized Wistar rats, rabbit biopsy, and human adipose tissue. These studies focused only on the static differences between cell-size distributions under different conditions. However, cross-sectional static cell-size distributions for a range of snapshots of animal development can be used to deduce the dynamics of adipose tissue growth, if we can appropriately analyze the snapshots with the help of mathematical modeling. Given present technical limitations, this may be the best available approach to an understanding of in vivo adipose tissue growth, although a recent experiment has observed lipid accumulation in lipid droplets of adipose cells.? To address genetic and dietary effects on the dynamic process of adipose tissue growth, we obtained cell-size distributions of epidydimal fat of obesity-resistant FVB/N (hereafter FVB) and obesity-prone C57Bl/6 (C57) mouse strains under standard chow and high-fat diets. The C57 mouse is the best characterized model of diet-induced obesity, and the FVB mouse is a preferable model for generating transgenic mice. These two commonly-used inbred mouse strains are genetically quite distant, and they have distinct metabolic phenotypes: Compared with FVB mice, C57 mice have low circulating triglyceride levels and increased triglyceride clearance; FVB mice are characterized by relatively higher hepatic insulin resistance, counter-regulatory response to hypoglycemia, and reduced glucose-stimulated insulin secretion; FVB mice are also known to be relatively lean since they appear to be less responsive to high-fat diet than C57 mice. However, the development of diet-induced obesity in these two strains has not been formally compared. In this study, we developed a mathematical model describing the change of the cell-size distribution as a function of the epidydimal fat pad mass to analyze quantitatively the dynamic characteristics that depend on genetics and/or diet. The model of adipose tissue growth describes how many new cells are formed, how each cell grows depending on its size, and how lipid turnover leads to size fluctuations that cause a spreading in the cell-size distribution. As the epidydimal fat pad mass increases, the cell-size distribution changes in a systematic manner depending on both genetics and diet. Comparing experimental results with the theoretical growth model, we found that hypertrophy is strongly correlated with diet. Hyperplasia, on the other hand, is initially dependent on genetics but is also affected by diet.? Our central finding is that hyperplasia and hypertrophy of adipose cells in the epidydimal fat pad is a function of the fat pad mass, even though it may take individual animals different time periods to reach a given fat pad mass. Therefore, adipose tissue growth, represented as changes of the cell-size distribution, can be systematically modeled as a growth process with respect to fat pad mass increase; this may reflect a correlation between fat pad mass and the secretion of adipokines and other signaling molecules controlling adipose tissue growth. Accordingly, it should be noted that the rates in our model are not the usual rates per unit time increase but rates per unit mass increase. Thus, several rates in the model had larger values for animals on a chow diet than for those on a high-fat diet. However, if these rates are converted to the usual rates per unit time increase, they had larger values for the high-fat diet, because it takes less time to effect a unit increase inthe fat pad mass from larger, and more numerous, cells on a high-fat diet than to effect an increase of the same magnitude from smaller, and fewer, cells on a chow diet.? It has been suggested that when obesity progresses, hypertrophy of adipose cells occurs early, and then triggers hyperplasia. Our study showed that new cell recruitment increases exponentially as fat pad mass increases. Hypertrophy of adipose cells is the main contributor to fat pad mass increase, whereas hyperplasia does not contribute much to this increase because it occurs in small cells that have a much smaller volume of fat stored. Therefore, our model naturally embodies the concept that hyperplasia is affected by the hypertrophic growth of cells. On the other hand, there have been reports that hyperplasia of adipose cells occurs only at early development stages; hence, there is no new cell recruitment at late stages even under obesogenic conditions. It may be the case that the age of the animals in our study (6 weeks old) allows the occurrence of hyperplasia.? The model developed here may give microscopic insights into the size-dependent growth of adipose cells that cannot be addressed by static cross-sectional studies. We found specific properties of the size-dependent cell growth: the lower critical size, initializing lipid accumulation with enough lipid transporters, did not depend on diets in two mouse strains, whereas the upper critical size, limiting the cell growth with an extraordinary size, was enlarged on a high-fat diet. It may be of interest to see if these results can be generalized to other strains and organisms.? In the tissue growth model, we included the recruitment of new cells and the growth of existing cells, but not the death of old cells, because the model was consistent with the data without the apoptosis of adipose cells. One recent study has reported that there is a turnover of human fat cells on a ten year time scale. Cell turnover may be irrelevant within the twelve week period of our study, but the model may need enhancement when it is applied to other fat depots that have functional differences, and to other species such as human.
| Jo, Junghyo; Gavrilova, Oksana; Pack, Stephanie et al. (2009) Hypertrophy and/or Hyperplasia: Dynamics of Adipose Tissue Growth. PLoS Comput Biol 5:e1000324 |
