1 Division of Molecular Genetics, Department of Pediatrics, Columbia University College of Physicians and Surgeons, New York, New York 10032
2 Max-Planck-Institut, fuer physiologische und klinische Forschung, W. G. Kerckhoff-Institut, D-61231 Bad Nauheim, Germany
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
energy metabolism; obesity; sympathetic nervous system; leptin receptor; adipose tissue
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
We have previously shown that leptin mRNA levels in inguinal WAT, body fat mass, and plasma-LEP per unit of fat mass are increased in 10-day-old Zucker rat pups in proportion to gene dosage for the functionally inactive leptin receptor, Leprfa (54). These effects of Leprfa are independent of plasma insulin concentrations (54), suggesting that leptin expression in adipose tissue is subject to negative feedback regulation that is mediated by the leptin receptor. Similar mutation dose-dependent increases in body fat content and plasma-LEP relative to fat mass were also observed in adult mice segregating for Leprdb (8). These effects of gene dosage for Lepr mutations could be conveyed via autocrine and/or CNS-mediated effects of leptin action (5, 12, 38, 56).
Decreased sympathetic nervous system (SNS) tone is a major characteristic of animals with defects in the leptin axis (10, 53). SNS activity is an important regulator of leptin gene expression in adipocytes. Sympathetic activation by cold exposure, or in vivo or in vitro administration of ß3-adrenergic receptor agonists, inhibits leptin mRNA expression in BAT and WAT in rodents (13, 17, 29, 48). Thus SNS activity could play an important role in mediating leptins feedback regulation of its own expression in adipose tissue. Leptin also directly inhibits its own mRNA levels in adipose tissue (56) and exerts direct effects on lipolysis and glucose utilization in adipocytes in vitro (19, 40, 42). Thus adipocyte leptin receptor (LEPR) may mediate direct effects on Lep expression and/or indirect effects through effects on metabolite flux in adipocytes (19, 29, 42, 52, 56).
BAT is an important thermogenic organ not only in cold-adapted rodents but also in newborns of most species, including human infants. In rats it constitutes the only adipose tissue visible at birth (14). Both BAT and WAT contain triglycerides and express many common adipocyte-specific genes, such as aP2, Lep, PPAR- (29, 42, 46). However, BAT differs from WAT most significantly in its thermogenic capacity and extensive sympathetic innervation (2, 3, 41). These differences in function, and in the extent of SNS innervation between BAT and WAT, make them ideal tissues in which to examine the role of SNS activity in conveying LEPR-mediated suppression of Lep gene expression in adipocytes. Although Lep gene expression is very low in BAT of adult animals (9), its expression is more abundant in BAT of newborn animals and suckling pups (14, 15).
Artificial rearing of rat pups under thermoneutral conditions with or without norepinephrine (NE) administration has been used to assess SNS-mediated effects on body fat content in fa/fa and +/fa rat pups (20, 21, 26). Under normal rearing conditions (pups suckled by dam), pups are intermittently exposed to cold loads whenever the mother leaves the nest, resulting in intermittent activation of the SNS (35, 36). Activation of thermoregulatory thermogenesis under such circumstances is much smaller in fa/fa pups than in +/fa and +/+ pups (32, 36). Artificial rearing under thermoneutral conditions prevents such transient stimulation of sympathetic tone, leading to a decreased but equal sympathetic tone in fa/fa and +/fa pups as indicated by identical rate of energy expenditure (20, 26). Artificial rearing under thermoneutral conditions with synthetic milk supplemented with NE, an SNS effector, mimics continuous activation of SNS in both fa/fa and +/fa pups and results in the same increased "effective SNS tone" in both fa/fa and +/fa pups (21). In addition, artificial rearing allows experimental control of caloric intake, enabling manipulation of metabolic flux and energy storage independently from energy expenditure in pups.
To investigate the role of SNS activity and metabolic flux and/or energy storage in mediating the feedback regulation of leptin on its own gene expression in adipocytes, we examined the effects of Leprfa gene dosage on leptin mRNA levels in BAT and WAT and plasma-LEP in 10-day-old rat pups under normal rearing conditions (mother rearing) and various artificial rearing conditions that affect SNS activity and caloric balance. In this model, which permits their experimental isolation, we compared the effects of differences in SNS activity, caloric balance, and body fat content on Lep mRNA expression in BAT and WAT and assessed their respective contributions to circulating leptin concentrations.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Definition and calculation of metabolic parameters.
1) Total daily energy expenditure (MR) was determined using a conversion factor of 20 kJ/liter O2 (26, 32). For those pups in which only Tc was recorded, MR during the final 24-h of life was extrapolated using regressions of MR against Tc of their littermates based on the close correlation between these two parameters in 10-day-old pups (33, 44, 45). 2) MS-MR is defined as total daily energy expenditure/body mass; MS-MR for artificially reared 10-day-old pups under thermoneutral conditions is 89 W/kg (20, 21, 26). 3) Caloric balance is defined as the difference between milk calories fed and energy expenditure (MR) during the last 24 h of life.
Specific conditions for artificial rearing.
1) To artificially rear pups under cold loads similar to those experienced during normal rearing conditions, ambient temperature for litters 8 and 9 (n = 4 fa/fa, 7 +/fa, and 6 +/+) was adjusted so that the Tc of the +/fa littermates was in the range observed in mother-reared +/+ and +/fa pups (31, 35, 36, 39), causing MS-MR during the last 24 h of life to be about 70% above that of thermoneutrally reared pups of this age (20, 21, 26). All pups in these litters were fed at the maximal feeding rate compatible with the stomach emptying of the hypothermic fa/fa pups. One of these cold-reared litters (litter 8) had higher fat-free dry mass (FFDM) than predicted, indicating that the pups might have been a day older than expected. The data from this litter were included in all scatter plots (Figs. 2 and 4), but not in any of the statistical analyses. 2) For thermoneutral rearing without NE treatment (litters 1013, n = 23 fa/fa, 13 +/fa, and 5 +/+), the ambient temperature was adjusted to keep Tc slightly above 37°C, i.e., at the temperature at which the average daily MR was at its thermoneutral value (21, 31). Feeding rate was maintained at a rate similar to that in the cold-reared litters. 3) For thermoneutral rearing with NE treatment (litters 1416, n = 13 fa/fa, and 13 +/fa), oral NE treatment under thermoneutral conditions was carried out exactly as previously described (21). In short, pups were started on artificial rearing with milk without NE on postnatal day 4. Starting at 5 days of age, pups were fed milk containing NE at concentrations that were identical for all pups in one litter. The dose of NE was adjusted twice daily to gradually increase MS-MR from its thermoneutral level on postnatal day 4 to about 100% above this level on postnatal day 10 or 11. Doses between 250 and 1,000 µg·kg-1·h-1 of free base were needed to obtain this effect. Ambient temperature was gradually decreased as necessary to maintain Tc between 37 and 39°C. To distinguish between the effects of variation in MR and variation in caloric balance and/or body fat content on leptin gene expression, we varied the milk feeding rates among the three NE-treated litters while keeping their MS-MR at similarly high levels. In two of these litters (litters 14 and 15), milk feeding rate throughout artificial rearing was maintained at the highest level compatible with the rate of stomach emptying, producing animals with a higher final body fat content (FatHI, see Figs. 2 and 3). In litter 16, milk feeding rate was kept at the lowest rate compatible with a normal growth rate (an average weight gain of 0.7 g/d), producing animals with a lower final body fat content (FatLO, see Fig. 2 and 3).
|
|
Determination of genotype at Lepr.
Genotype was determined by the presence of a novel MspI site introduced by the Leprfa mutation. Genomic DNA isolated from tail clips was amplified by PCR using primers flanking the Leprfa mutation followed by MspI restriction digestion of the PCR products and gel electrophoresis analysis (7).
Determination of plasma-LEP.
Plasma-LEP were determined with a commercial murine leptin radioimmunoassay (RIA) kit (Linco, St. Charles, MO) and corrected against repeated measurements of a rat leptin standard (Linco) within the physiological range. In our hands, this method yielded more reproducible results, particularly in the lower range of plasma-LEP, than commercially available rat leptin RIA kits. Two or three aliquots of each plasma sample were assayed. The intra-assay and interassay variation was <5% and <10%, respectively.
Quantitative RT-PCR.
Total RNA from adipose tissue was extracted using TRIzol reagent (GIBCO; Life Technologies, Bethesda, MD). About 15 µg of total RNA was reverse-transcribed into cDNA using random hexamer and Moloney murine leukemia virus reverse transcriptase. Complementary DNA was amplified by PCR using specific primers for Lep, aP2, or Ucp1 in combination with primers for ß-actin (as an internal control) as previously described (54). Primers for Lep were 5'-TGACACCAAAACCCTCATCA-3' (forward) and 5'-AGCCCAGGAATGAAGTCCA-3' (reverse), for Ucp1 were 5'-GGTGAGTTCGACAACTTCCG-3' (forward) and 5'-GGTGATGGTCCCTAAGACACC-3' (reverse), for aP2 were 5'-GACGACAGGAAGGTGAAGAGC-3' (forward) and 5'-GCCTTTCATAACACACATTCCACC-3' (reverse), and for ß-actin were 5'-AGGCCCAGAGCAAGAGAG-3' (forward) and 5'-GGGTGTTGAAGGTCTCAAAC-3' (reverse). [32P]dCTP-labeled PCR products were visualized and quantified using a PhosphoImager (Molecular Dynamics). Message RNA levels for Lep, aP2, and Ucp1 were expressed as a ratio to ß-actin mRNA to normalize for initial RNA input.
Statistical analyses.
Statistical analysis was performed using Statistica 6.0 or Sigma Stat. Regression analysis has advantages over ANOVA in analytic strategies designed to take account of interlitter variability in growth, body composition, and other phenotypes as they relate to treatments applied to the various litters (24). For regression analyses, plasma-LEP and leptin mRNA levels in BAT and WAT were log transformed to normalize their respective distributions. In preliminary forward stepwise regression analyses, a best set of independent variables were selected for their ability to generate the highest predictive values (R2) for the dependent variable. Individual variables with F values <1 were excluded from the respective regression analysis. Gender was not included in the analyses as an independent variable, because F values for gender in all analyses reported here were <1.3 in primary analyses of the data. The percent of variability accounted for and the standardized correlation coefficient (ß) were used to characterize the effects of various independent variables on leptin mRNA levels in BAT and WAT and plasma-LEP.
Effects of genotype at Lepr on plasma-LEP and leptin mRNA expression in BAT and WAT in mother-reared pups were also evaluated by two-way ANOVA with litter and genotype as grouping factors to allow for the well-known litter effects on growth of suckling-age pups (28, 39, 49). Results are presented as least square means ± SE. To allow inclusion of litters containing only two of three possible genotypes at Lepr, differences between fa/fa and +/fa, and between +/fa and +/+, were examined in separate ANOVAs. In addition, ANCOVA using litter and genotype at Lepr as grouping factors and body fat mass as a covariate was performed to evaluate the effect of genotype at Lepr on plasma-LEP adjusted for body fat mass.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Effects of experimental manipulations of sympathetic nervous activity and caloric balance on body fat content and plasma-LEP.
To identify the mechanisms for the control of leptin mRNA expression in BAT and WAT and their respective contributions to circulating plasma-LEP, variations in SNS activity, body fat content, and metabolic flux were created by experimentally manipulating ambient temperature and the rate of milk infusion in combination with NE administration during the artificial rearing of rat pups from postnatal day 4 or 8 to day 10/11. Figure 2B shows the MS-MR during the last 24 h of life of artificially reared pups. As previously reported (26), rearing pups under moderate cold loads increased MS-MR in +/fa and +/+ pups relative to pups reared under thermoneutral conditions, while MS-MR in fa/fa pups remained close to that observed under thermoneutral conditions, resulting in hypothermia in the fa/fa pups. Rearing pups of all genotypes under thermoneutral conditions equalized MS-MR at very low levels. Adding NE to the infused synthetic milk increased MS-MR to high but equal levels among fa/fa and +/fa or +/+ pups. MS-MR in NE-treated pups in the final 24 h of the experiment was about 100% above that observed in pups reared under thermoneutral conditions without NE treatment. MS-MR has been previously shown to be highly correlated with sympathetically mediated activation of BAT in pups at this age (16, 21, 22, 33) and was used as a surrogate for SNS activity in these pups. The manipulation of feeding rates in the NE-treated litters produced two groups of pups with similarly high SNS activity but markedly different body fat content and caloric balance (litters 14 and 15 vs. litter 16). These nutritional manipulations permit isolation of the effects of SNS activity from those of body fat content and/or caloric balance on leptin gene expression.
Because artificially reared pups were smaller than mother-reared pups, percentage of body fat content rather than the absolute fat mass was used to compare their adiposity to that of mother-reared animals (Fig. 2C). When pups were artificially reared under moderate cold loads (litters 8 and 9) similar to those experienced by mother-reared pups (31, 35), the Leprfa-dependent differences in body fat content and plasma-LEP were at least as pronounced as in mother-reared pups (Fig. 2, C and D), demonstrating that the artificial rearing procedure per se does not eliminate the abnormally high plasma-LEP of fa/fa pups.
Compared to cold rearing, rearing under thermoneutral conditions for 2 days (from postnatal day 8 to 10) resulted in a striking (more than 3-fold) decrease of plasma-LEP in fa/fa pups but had little effect on plasma-LEP in +/fa and +/+ pups (litters 10 and 11). This result was unexpected, because activation of the SNS has previously been shown to suppress plasma-LEP in lean mice and rats (17, 29, 48). Thus minimizing SNS activity by thermoneutral rearing had been expected to increase leptin expression in WAT in +/+ and +/fa pups instead of decreasing it in fa/fa pups. When pups were reared under thermoneutral conditions for 6 days (from postnatal day 4 to day 10/11), body fat content and plasma-LEP were indistinguishable in +/fa and fa/fa pups (litters 12 and 13). Proportional to the increase in body fat content after 6 days under thermoneutral conditions, the plasma-LEP of both fa/fa and +/fa were higher than in the pups reared under thermoneutral conditions for only 2 days. These results suggest that artificial rearing under thermoneutral conditions equalized energy expenditure between fa/fa and +/fa pups and eliminated the factor(s) that caused increased plasma-LEP in mother-reared or artificially reared fa/fa pups under moderate cold loads.
The NE treatment decreased plasma-LEP to very low, but similar, levels in both fa/fa and +/fa pups (litters 1416), even though the body fat contents of both fa/fa and +/fa pups in litters 14 and 15 were similar to those of the fa/fa pups reared under cold load. These results demonstrated that moderate doses of NE were sufficient to suppress plasma-LEP to similarly low levels in both fa/fa and +/fa pups without concomitantly decreasing body fat content.
Effects of experimental manipulations of SNS activity and caloric balance on leptin expression in WAT and BAT.
We examined leptin mRNA levels in WAT and BAT of pups in which caloric balance and SNS activity had been manipulated to introduce variations in metabolic flux and/or energy storage that were independent of altered SNS activity. In the pups artificially reared under moderate cold loads, WAT-Lep mRNA showed a similar Leprfa dose-dependent relationship among fa/fa, +/fa and +/+ pups as was seen in mother-reared pups, although the sample size for this group of pups was too small to yield statistically significant differences between the genotypes (Fig. 3A). Minimizing SNS activity by thermoneutral rearing for either 2 or 6 days equalized WAT-Lep mRNA in the +/fa and fa/fa pups by decreasing Lep expression in fa/fa pups. Lower SNS activity would be expected to increase Lep expression in pups of all genotypes (13, 48). The unexpected result here suggests that SNS activity per se is not an important determinant of Lep expression in WAT at this age (see below). Increasing SNS activity by NE treatment was not sufficient to decrease or equalize WAT-Lep mRNA in fa/fa and +/fa pups from litters with high body fat content (FatHI, Figs. 2 and 3). WAT-Lep mRNA was nearly twofold higher in the fa/fa than the +/fa pups (P = 0.055), despite their similar body fat contents (Fig. 2C). In comparison, decreasing body fat content and caloric balance by reducing feeding rate in the NE-treated litter (FatLO) markedly suppressed WAT-Lep mRNA levels in both fa/fa and +/fa pups, implying that energy flux and/or fat content play an important role in regulating leptin expression in WAT of these pups.
The effects of rearing conditions on leptin gene expression in BAT of the same artificially reared animals were very different from those in WAT observed above. BAT-Lep mRNA in cold-reared pups was much lower than that in mother-reared pups (Fig. 3B). But, as in mother-reared pups, BAT-Lep mRNA in fa/fa pups was severalfold higher than that of their +/+ and +/fa littermates, whose levels were not different from each other. Thermoneutral rearing pups for 2 or 6 days increased BAT-Lep mRNA in both +/fa and fa/fa pups in a time-dependent manner to much higher levels, relative to pups either mother-reared or artificially reared under moderate cold loads. Thermoneutral rearing also completely eliminated the differences in BAT-Lep mRNA between +/fa and fa/fa pups. Thus, in contrast to WAT-Lep mRNA, BAT-Lep mRNA showed the expected increase when SNS activity was decreased by rearing under thermoneutral conditions. The potent impact of SNS activity on BAT-Lep mRNA is also emphasized by the dramatic suppression and equalization of BAT-Lep mRNA to very low levels in NE-treated +/fa and fa/fa pups, even in litters in which increased rate of milk infusion had prevented a decrease of body fat content (FatHI, Figs. 2 and 3).
To further characterize the effects of SNS activity, metabolic flux, and/or energy storage and genotype at Lepr on leptin mRNA expression in adipose tissues, we performed multiple regression analyses using BAT-Lep mRNA and WAT-Lep mRNA as dependent variables, and MR (as a surrogate for SNS activity), caloric balance and body fat mass (as surrogates for metabolic flux and energy storage), FFDM, and genotype at Lepr locus as independent variables. FFDM, which is a more precise measure of lean body mass than wet body mass, was included in these analyses to adjust for differences in the metabolic size among the pups. BAT-Lep mRNA showed a strong inverse correlation with MR (ß = -0.78, P < 0.001) and a moderate positive correlation with body fat mass (ß = 0.37, P < 0.001) (total R2 = 0.93) (Table 1). In stepwise regression analysis, MR and fat mass accounted for 87% and 5% of the variation in BAT-Lep mRNA, respectively. WAT-Lep mRNA, in contrast, was mainly influenced by caloric balance (ß = 0.55, P < 0.001) and fat mass (ß = 0.39, P < 0.01)(total R2 = 0.53) (Table 1). In stepwise regression analysis, caloric balance and fat mass accounted for 40% and 6% of the variation of WAT-Lep mRNA, respectively. MR also had a weak positive impact on WAT-Lep mRNA, accounting for 5% of the variation (ß = 0.36, P < 0.01), in contradistinction to the negative impact seen in BAT-Lep and in previous studies of WAT (17, 29, 48). This unexpected result could be attributed to the very low WAT-Lep mRNA in fa/fa pups reared under thermoneutral conditions without NE treatment. Indeed, excluding these pups in the regression analysis eliminated the correlation of MR with WAT-Lep mRNA.
|
Surprisingly, copy number of Leprfa had no significant effect on either BAT-Lep mRNA or WAT-Lep mRNA when MR, body fat mass, and caloric balance were considered in the regression analysis, suggesting that the Leprfa may not have a direct (autocrine) effect on leptin mRNA expression in adipose tissues of 10-day-old rat pups. Thus the observed effects of Lepr-mediated feedback suppression on leptin mRNA expression in mother-reared pups appear to be due mainly to LEPR-mediated effects on SNS activity in BAT and to effects on energy flux and/or energy storage in WAT.
The role of WAT and BAT leptin mRNA levels and associated physiological parameters in determining plasma-LEP.
To assess the relative contributions of rates of leptin expression in BAT and WAT to the circulating leptin concentrations in artificially reared pups, we performed multiple regression analyses with plasma-LEP as the dependent variable and BAT-Lep and WAT-Lep mRNA levels and genotype at Lepr as independent variables (Table 1). It was noted that all of the fa/fa pups reared under cold loads, and 2 of the 34 fa/fa pups reared under thermoneutral conditions, had plasma-LEP (>10 ng/ml) much higher than the other artificially reared pups. Multiple regression analysis with all artificially reared pups gave a lower total R2 = 0.61 than in the analysis when pups with plasma-LEP > 10 ng/ml were excluded (R2 = 0.70). Under the latter analysis paradigm, BAT-Lep mRNA accounts for 60% of the variation in plasma-LEP, whereas WAT-Lep mRNA accounts for only 3% of the variation, consistent with the results obtained in mother-reared pups.
Genotype at Lepr accounted for 7% of the variation in plasma-LEP, suggesting that the functional integrity of LEPR has effects on plasma-LEP by mechanisms separate from its effects on leptin mRNA expression in BAT and WAT. Inclusion of MR (surrogate for SNS activity) as an additional independent variable in the regression analysis did not increase the total R2 value (data not shown). Because of the significant effect on Lep expression of genotype at Lepr, we also evaluated +/fa and fa/fa pups in separate regression analyses. The correlation of BAT-Lep mRNA or WAT-Lep mRNA with plasma-LEP was higher in fa/fa pups (ß = 0.84 and 0.31, respectively) than in +/fa pups (ß = 0.73 and 0.13, respectively). However, the analyses confirmed the predominant effects of BAT, relative to WAT, on plasma-LEP in both fa/fa and +/fa pups.
To directly assess the relationship between leptin mRNA expression in adipose tissues and plasma-LEP in pups under the various rearing conditions, we performed simple correlation analysis of WAT-Lep mRNA or BAT-Lep mRNA with plasma-LEP (Fig. 4, A and B). These analyses support the conclusions based upon the multiple regression analysis described above: plasma-LEP values were very strongly correlated with BAT-Lep mRNA (r = 0.78, P < 0.001) (Fig. 4B) but less strongly with WAT-Lep mRNA (r = 0.39, P < 0.01) (Fig. 4A). The least square means of plasma-LEP, WAT-Lep mRNA, and BAT-Lep mRNA from mother-reared pups (after controlling for litter effects, oversized solid symbols in Fig. 4) are included to permit a direct comparison between mother-reared and artificially reared pups. The means for mother-reared +/+ and +/fa pups were within the range of artificially reared +/+ and +/fa pups. However, the mean plasma-LEP of the mother-reared fa/fa pups was above 10 ng/ml, similar to those found in the fa/fa pups artificially reared under cold loads (large solid squares with dot and cross, in Fig. 4). In contrast, only 2 of 34 fa/fa pups reared under thermoneutral conditions had plasma-LEP >10 ng/ml (open square with cross, Fig. 4). Thus it appears that cold exposure increased plasma-LEP in both mother-reared and artificially reared fa/fa pups. But such an effect was not observed in the lean (+/fa and +/+) pups.
We also analyzed MR, caloric balance, body fat content, and genotype at Lepr for their effects on plasma-LEP in artificially reared pups by multiple regression analysis (Table 1). As expected, plasma-LEP was inversely related to MR (ß = -0.58, P < 0.001) and positively influenced by body fat content (ß = 0.50, P < 0.001). Caloric balance and FFDM also showed a weak, negative correlation with plasma-LEP. Total R2 was 0.70. No additional effect of genotype at Lepr on plasma-LEP was detected, suggesting that the effect of the Leprfa gene on plasma-LEP observed in mother-reared pups was mediated mostly through the effects of Leprfa gene on SNS activity, metabolic flux, and/or energy storage.
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The failure of exogenous NE to suppress Lep expression in WAT of the pups in NE-treated FatHI group (Fig. 3) was surprising. Previous reports have indicated that cold stress or administration of catecholamines or ß3-adrenergic receptor agonists suppresses leptin gene expression in WAT of adult lean mice, although the effects of these manipulations on leptin expression in WAT were much less potent than in BAT (29, 48). Since most reported effects of SNS activation or administration of ß-adrenergic agonist on leptin gene expression in adult animals were obtained by experiments without appropriate control of substrate flux in adipocytes (13, 17, 29, 48), it is possible that the sympathetic-mediated effects on leptin gene expression are secondary to sympathetic-mediated lipolysis in adipocytes. This idea is consistent with the fact that the suppression of leptin mRNA levels in adipocytes by catecholamines appears to be coupled to the synthesis of cAMP, which also stimulates lipolysis (13, 23, 43). Changes in fatty acids flux due to SNS-mediated lipolysis in adipocytes may influence leptin gene expression by effects on the hexosamine biosynthetic pathway (51, 52) or by other mechanisms. Consistent with our results in suckling pups, Moinat et al. (29) have also shown that in adult rats, food restriction decreases leptin mRNA levels in WAT to a much greater extent than either cold-stimulation or administration of ß3-adrenergic agonist. Based on these results, Moinat et al. (29) suggested that leptin gene expression in WAT might be more closely associated with alterations in cell lipid content and/or substrate flux than with changes in SNS activity per se. Thus our results in suckling pups agree with those of Moinat et al. (29) in adult rats with regard to the relatively higher sensitivity of leptin gene expression in WAT to changes in substrate flux than to changes in SNS activity per se, even though the relative size of BAT and WAT and their contributions to the circulating leptin concentration differ significantly during the suckling period and adult life. However, it is important to note that our results do not exclude the possibility of regulation of leptin gene expression in WAT by SNS activity in either pups or adult rats, but only emphasize the relative importance of substrate flux in control of leptin gene expression in WAT. The more potent effects of metabolic flux into adipocytes/energy storage may mask the relatively weak effects of enhanced SNS activity on Lep expression in WAT of pups in NE-treated FatHI group (Fig. 3). It is also worth noting that the total R2 for SNS activity, body fat mass, and caloric balance in predicting WAT Lep mRNA levels was relative low (0.53), compared with 0.93 in BAT, suggesting that factors other than SNS activity or metabolic flux/energy storage may be involved in mediating the regulation of WAT-Lep expression in the pups.
Contributions of leptin expression in BAT and WAT to plasma-LEP.
The higher degree of correlation between plasma-LEP and leptin mRNA expression in BAT than in WAT in both mother-reared and artificially reared pups suggests that BAT is a more significant source of circulating leptin than WAT in 10-day-old rat pups. Since transcription rates of leptin (relative to actin) were severalfold lower in BAT than WAT in mother-reared and artificially reared pups under cold load or NE treatment, the greater mass of BAT (vs. WAT) at this age may partly account for the higher total leptin production by BAT than WAT. Although the relative contribution of WAT to circulating leptin levels, compared with that of BAT, is apparently lower in these 10-day-old pups, the contribution of the WAT depot increases with age as WAT mass increases. In adult animals, plasma-LEP correlates highly with both WAT-Lep mRNA and the mass of WAT (18, 25, 37).
Interestingly, although BAT-Lep mRNA expression rate was more closely correlated with plasma-LEP, plasma-LEP was not always correlated with BAT-Lep mRNA under all artificial rearing conditions. For example, the dramatic increase of BAT-Lep mRNA in both fa/fa and +/fa pups reared under thermoneutral conditions without NE for 6 days was not accompanied by high plasma-LEP. On the other hand, the very high plasma-LEP (>10 µg/ml) levels that were observed in fa/fa pups artificially reared under cold loads were associated with WAT-Lep mRNA and BAT-Lep mRNA that were either similar or lower than those seen in pups reared under thermoneutral condition without NE treatment for 6 days. These results suggest that plasma-LEP in these pups may also be regulated by mechanisms that are not associated with leptin production in BAT and WAT, such as the clearance of the circulating leptin. The kidney has been shown to be a major site for clearance of circulating leptin by glomerular filtration (11, 50). Hypothermia decreases renal blood flow and glomerular filtration rate by non-SNS-mediated mechanisms (4, 6). Thus decreased clearance of circulating leptin may contribute to the marked increase of plasma-LEP in hypothermic relative to euthermic fa/fa pups.
Mechanisms of Lepr-mediated feedback suppression of leptin gene expression.
The importance of leptin signaling in energy homeostasis during the early postnatal life of rat pups is clearly demonstrated by the onset of increased fat deposition in pups that are either heterozygous or homozygous for the mutant leptin receptor (Leprfa) (28, 39, 49, 54). Leptin administration also alters the circadian rhythm of oxygen consumption and reduces time spent in a torpid state in suckling pups (44). The Leprfa gene dose-dependent increases in BAT and WAT Lep mRNA levels, as well as plasma-LEP, suggest that leptin signaling plays a role, directly or indirectly, in controlling leptin expression at this age. The increased leptin mRNA in BAT and WAT of fa/fa pups was not associated with a general increase of gene transcription rates, since BAT Ucp1 and aP2 mRNA levels were not significantly different between mother-reared fa/fa and +/fa pups (data not shown). The lack of effect of Leprfa on Ucp1 mRNA levels in the BAT of the 10-day-old fa/fa pups is consistent with a previous report in which decreased thermoregulatory thermogenesis in 10-day-old fa/fa rats was shown to be due to decreased GDP binding capacity of UCP1 protein, rather than to decreased Ucp1 mRNA or protein levels (1). However, significant decreases in BAT Ucp1 mRNA and protein levels have been found in older fa/fa rats relative to lean littermates (1).
The finding that BAT-Lep mRNA levels were not different between fa/fa and +/fa pups when their SNS activity was equalized by thermoneutral rearing suggests that SNS constitutes the major mechanism for the LEPR-mediated feedback suppression on leptin gene expression in BAT of pups at this age. The apparent insensitivity of WAT leptin mRNA levels to NE treatment in both fa/fa and +/fa pups suggests that decreased SNS activity is not the primary mechanism for the increased WAT leptin mRNA levels in fa/fa mother-reared pups. The marked difference of WAT-Lep mRNA between NE-treated pups with high (FatHI) and low (FatLO) body fat content and caloric balance emphasizes the importance of metabolic flux/energy storage in regulating leptin expression in WAT (Fig. 3).
Thermoneutral rearing equalized WAT-Lep mRNA in fa/fa and +/fa pups by decreasing WAT leptin expression in fa/fa pups, rather than increasing WAT-Lep mRNA in +/fa pups, as one would have predicted based on the effects of thermoneutral rearing on BAT-Lep mRNA and known effects of SNS activity on leptin mRNA expression in WAT of adult animals (17, 29, 48). One possible factor responsible for the higher WAT-Lep mRNA levels in fa/fa than +/fa pups reared in a colder ambient temperature (either mother-reared or artificially reared) may be the general physiological stress induced by hypothermia in fa/fa pups. Rearing pups under thermoneutral conditions prevents hypothermia and the increase of WAT-Lep mRNA in fa/fa. Glucocorticoids, which increase leptin expression in adipocytes, might mediate the increased leptin expression in WAT of fa/fa pups reared under relatively cold conditions (43). Experimental manipulation of glucocorticoid status would address this possibility. Alternatively, the relative increases of metabolic flux toward WAT due to the decreased energy expenditure for thermogenesis in BAT of fa/fa pups may be responsible for the higher WAT-Lep mRNA in fa/fa than +/fa pups under cold loads.
No significant effect of genotype at Lepr on leptin mRNA levels in BAT and WAT was detected in artificially reared pups after controlling for SNS activity, caloric balance, and body fat content. This result suggests that the effects of Leprfa gene dosage on leptin mRNA levels in BAT and WAT observed in mother-reared pups are largely mediated by Leprfa effects on SNS activity and metabolic flux/energy storage. The effects of Lepr on SNS and metabolic flux/energy storage may be mediated mainly by LEPR in the hypothalamus (5, 47). However, leptins effects on metabolic flux in WAT and BAT conveyed directly by leptin receptors on adipocytes may also be involved in the feedback regulation of leptin on its own gene expression (27, 40, 42, 56).
In summary, the results reported here show: 1) Leptin mRNA expressions in BAT and WAT of 10-day-old rat pups are differently regulated; BAT-Lep mRNA is strongly suppressed by SNS activity, whereas WAT-Lep mRNA is primarily influenced by adipocyte metabolic flux/energy storage. 2) Although BAT-Lep mRNA levels are severalfold lower than WAT-Lep mRNA levels, plasma-LEP levels in pups of this age are predominantly related to BAT-Lep mRNA expression rates. 3) Leptin expression in both BAT and WAT is increased in fa/fa pups compared with +/fa and +/+ pups. Decreased SNS activity is mainly responsible for the increased leptin mRNA level in BAT in fa/fa pups, but increased metabolic flux/energy storage is at least partly responsible for the elevation of leptin mRNA level in WAT of fa/fa pups at this age.
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
FOOTNOTES |
---|
Address for reprint requests and other correspondence: Y. Zhang, Division of Molecular Genetics, Columbia Univ., Russ Berrie Pavilion, 1150 St. Nicholas Ave., New York, NY 10032 (E-mail: yz84{at}columbia.edu).
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|