Aging, depot origin, and preadipocyte gene expression

Abstract

Fat distribution changes with aging. Inherent changes in fat cell progenitors may contribute because fat cells turn over throughout life. To define mechanisms, gene expression was profiled in preadipocytes cultured from epididymal and perirenal depots of young and old rats. 8.4% of probe sets differed significantly between depots, particularly developmental genes. Only 0.02% differed with aging, despite using less stringent criteria than for comparing depots. Twenty-five genes selected based on fold change with aging were analyzed in preadipocytes from additional young, middle-aged, and old animals by polymerase chain reaction. Thirteen changed significantly with aging, 13 among depots, and 9 with both. Genes involved in inflammation, stress, and differentiation changed with aging, as occurs in fat tissue. Age-related changes were greater in perirenal than epididymal preadipocytes, consistent with larger declines in replication and adipogenesis in perirenal preadipocytes. Thus, age-related changes in preadipocyte gene expression differ among depots, potentially contributing to fat redistribution and dysfunction.

Figure 1.

Preadipocyte expression profiles differ extensively among fat depots and less prominently with aging. A). Regional variation in preadipocyte expression profiles. Nine hundred and twenty-one transcripts (of 10,983 probe sets detected) demonstrated both a twofold or greater difference in expression between undifferentiated perirenal compared with epididymal preadipocytes from 3- and 30-month-old rats and a depot-effect false discovery rate (FDR) less than 0.01 by analysis of variance (ANOVA). Data shown were z-score normalized and organized by hierarchical clustering, with each column representing a single animal. Full names, accession numbers, and hybridization intensities are in Supplementary Table 1. B). Stathmin-like 2 (Stmn-2) and tetranectin (Tna) increase with aging in preadipocytes. Only 3 out of 10,983 probe sets demonstrated significant age-dependent differences in expression in comparisons of preadipocytes isolated from 3-month compared with 30-month-old rats, despite use of less stringent criteria than those used to detect differences among fat depots (FDR < 0.08 by ANOVA and no fold-change criterion).

Figure 2.

Changes in gene expression with aging tend to be fat depot dependent. A). Genes most upregulated or downregulated between age groups were further tested by real-time polymerase chain reaction (PCR) in additional animals. RNA was isolated from perirenal and epididymal undifferentiated preadipocytes from young (3 months), middle-aged (17 months), and old (30 months) rats. Data shown are z-score-normalized and organized by hierarchical clustering. Each column represents data from a single animal (N = 7 animals in each age group). p values were computed by two-way fixed effects analysis of variance and adjusted using the false discovery rate method. Gene descriptions are in Table 4. B). Four additional genes relevant to fat tissue function or with distinctive profiles suggested from the array analyses were assayed by real-time PCR.

Figure 3.

Stmn-2, Mmp-3, and Mmp-12 increase more extensively with aging in perirenal than epididymal preadipocytes. Stathmin-like 2 (A), matrix metalloproteinase 3 (Mmp-3); (B), and Mmp-12 (C) mRNA levels in undifferentiated perirenal and epididymal preadipocytes cultured from young (3 months), middle-aged (17 months), and old (30 months) rats were assayed by real-time polymerase chain reaction (N = 7 animals in each group; *p < .05; Duncan’s multiple range test; data plotted from Figure 2A).

Figure 4.

Matrix metalloproteinase (Mmp) 3 and Mmp-12 proteins increase with aging in a depot-dependent manner. Lysates of undifferentiated perirenal and epididymal preadipocytes cultured from young (3 months), middle-aged (17 months), and old (30 months) rats were analyzed by Western blotting for Mmp3 (A) and Mmp12 (B). *p < .05 by analysis of variance; N = 7.