Abstract
Context: Lipin-1 functions in adipocyte triglyceride biosynthesis and in the regulation of gene expression, both of which may influence metabolic homeostasis.
Objective: Our objective was to determine whether variations in adipose tissue lipin-1 expression levels influence insulin sensitivity and gene expression in young healthy human subjects.
Design and Subjects: In 56 healthy young men (22.6 ± 3.2 yr; 26.4 ± 4.1 kg/m2) we determined insulin sensitivity by a euglycemic-hyperinsulinemic clamp, and whole body oxygen consumption and respiratory quotient by indirect calorimetry. We performed gene expression analysis in adipose tissue samples from human subjects and from lipin-1 transgenic mice using quantitative RT-PCR.
Results: In healthy young men, lipin-1 expression was positively correlated with insulin sensitivity (R2 = 0.22; P < 0.01), insulin-stimulated respiratory quotient (R2 = 0.16; P < 0.01), and maximal oxygen consumption during exercise (R2 = 0.16; P < 0.01). Lipin-1 mRNA levels were also correlated with expression of genes involved in lipid oxidation, uptake, and lipolysis, both in humans and in lipin-1 transgenic mice. The strongest correlation occurred between lipin-1 and peroxisome proliferator-activated receptor α (R2 = 0.74; P < 1 × 10−7), a nuclear receptor with a key role in fatty acid oxidation.
Conclusion: Lipin-1 expression levels in adipose tissue of healthy young subjects and in mice are correlated with a favorable metabolic profile and expression of fatty acid oxidation genes.
Obesity is a well-known risk factor for diabetes and cardiovascular disease, but the role of inter-individual variation in gene expression levels has not been well characterized. Lipin-1 mRNA levels in adipose tissue of healthy young subjects and in mice were correlated with insulin sensitivity, reduced fat cell size, and expression of fatty acid oxidation genes.
The prevalence of obesity is increasing in many parts of the world, and is considered a risk factor for diabetes and cardiovascular disease (1,2). It is clear that the development of obesity involves the interaction of many genetic factors with environmental factors such as diet and physical activity. Understanding the effect of genetic variation in the structure or expression of genes involved in lipid metabolism will provide a fuller understanding of the processes that lead to obesity and accompanying conditions, such as insulin resistance.
We previously identified lipin-1 (Lpin1), a gene that is highly expressed in adipose tissue and skeletal muscle, as the cause of lipodystrophy and insulin resistance in the fatty liver dystrophy mouse (3). Lipin-1-deficient mice fail to develop sc and visceral white adipose tissue depots, as well as interscapular brown adipose tissue (4,5). On the other hand, enhanced expression of lipin-1 in adipose tissue or skeletal muscle of transgenic mice leads to obesity (6). Interestingly, despite their greater fat mass, the adipose tissue-specific lipin-1 transgenic mice have improved insulin sensitivity.
Recent studies have revealed that lipin-1 has two cellular functions. First, lipin-1, as well as the related lipin-2 and lipin-3 proteins, act as Mg2+-dependent phosphatidate phosphatase type-1 (PAP1) enzymes, which catalyze a key step in the synthesis of glycerolipids (7,8,9). PAP1 converts phosphatidate to diacylglycerol, which serves as the direct precursor of triacylglycerol, and of the phospholipids phosphatidylcholine and phosphatidylethanolamine. Lipin-1 accounts for all of the PAP1 activity in white and brown adipose tissue, providing a mechanism for the lipodystrophy observed in lipin-1-deficient mice (8). Lipin-1 can also act as a transcriptional coactivator. Through direct interaction with peroxisome proliferator-activated receptor (PPAR) α and PPARγ coactivator-1 α, lipin-1 mediates enhanced expression of PPARα target genes involved in fatty acid oxidation in mouse liver (10). The significance of lipin-1 nuclear receptor coactivator activity in adipose tissue has not been investigated.
The strong effect of lipin-1 expression levels on adiposity and insulin sensitivity in mouse models raises the question of whether variations in lipin-1 expression levels in human adipose tissue likewise influence these traits. Thus far, studies performed on 19 dyslipidemic and 39 normal and obese subjects have shown correlations between lipin-1 expression levels in adipose tissue and indirect measures of insulin sensitivity, such as insulin and glucose levels or glucose tolerance (11,12). A strong association has also been observed between adipose tissue lipin-1 mRNA expression and basal and insulin-stimulated glucose transport in cultured human adipocytes (13). Although robust correlations were observed in these studies, the ability to make conclusions about the role of lipin-1 as a determinant of insulin sensitivity in these individuals was confounded by the presence of additional risk factors, such as hyperlipidemia and obesity, and by the use of surrogate measures of insulin sensitivity, such as glucose/insulin levels and glucose tolerance. Therefore, we set out to test the hypothesis that variations in lipin-1 expression levels in healthy young individuals are correlated with insulin sensitivity, as determined by euglycemic-hyperinsulinemic clamp. We also investigated the relationship between lipin-1 levels and adiposity, energy metabolism, and lipid metabolism gene expression in adipose tissue. Our results indicate that lipin-1 is correlated with insulin sensitivity and energy balance parameters in healthy young individuals, and may have a role in modulation of gene expression in adipose tissue, including effects on PPARα.
Subjects and Methods
Study population and design
After providing written informed consent, a cohort of 56 healthy young men, aged 22.6 ± 3.2 yr with a body mass index (BMI) of 26.4 ± 4.1 kg/m2, underwent physical examination, medical laboratory tests, and measurement of body fat by dual-energy x-ray absorptiometry (DXA). Subjects that were not weight stable were excluded from participation. Participants presented to the Pennington inpatient unit and ate a weight-maintaining diet (35% fat, 16% protein, and 49% carbohydrate) prepared by the metabolic kitchen. After 48 h of this diet, a euglycemic-hyperinsulinemic clamp was performed. Earlier studies describe this cohort in terms of skeletal muscle oxidative phosphorylation and metabolic flexibility (14,15,16,17). The present investigation only involves measures made before dietary intervention.
Euglycemic-hyperinsulinemic clamp
Glucose disposal rate (GDR) [mg/kg fat free mass (FFM)·min] during the insulin-stimulated state was measured by a euglycemic-hyperinsulinemic clamp (18). After an overnight fast, insulin (80 mIU/m2 BSA) was administered iv and glucose infused to maintain plasma glucose at 90 mg/dl for 2 h. GDR was expressed in terms of lean body mass, determined by DXA.
Maximal aerobic capacity (VO2max)
Maximal oxygen uptake was determined by a progressive treadmill test to exhaustion (19). Oxygen consumption (VO2) and CO2 production (VCO2) were measured continuously using a breath-by-breath indirect calorimeter (V-Max29 Series; SensorMedics, Yorba Linda, CA).
Body composition and adipose tissue mass
Body fat mass and lean body mass were measured on a Hologic dual-energy x-ray absorptiometer (QDR 4500; Hologic, Inc., Waltham, MA). Visceral adipose tissue mass was measured by multislice CT scanning as previously described (20).
Adipose tissue biopsy
After local anesthesia with lidocaine-bupivacaine, approximately 250 mg tissue was collected using Bergström needle biopsy from the lateral periumbilical sc adipose tissue.
Indirect calorimetry
After an overnight fast, the respiratory quotient (RQ) was measured at baseline by indirect calorimetry, and was measured again over 20 min during the euglycemic-hyperinsulinemic clamp using a DeltaTrac II indirect calorimeter (DATEX-Ohmeda, Helsinki, Finland). Oxidative and nonoxidative glucose disposal was calculated as described by Livesey and Elia (21).
Fat cell size
Fat cell size was determined as previously described (22). Briefly, adipose tissue was fixed in osmium tetrachloride/collidine-HCl, followed by disassociation via urea digestion. Cells were counted on a Multisizer-3 (Beckman Coulter, Fullerton, CA) using a 400-μm aperture (dynamic linear range 12–320 μm) and reported as the mean of all adipocytes more than 22.5 μm.
Laboratory measures
Fasting serum glucose and free fatty acids (FFA) were assayed on a Beckman Synchron CX7 (Beckman Coulter, Brea, CA). Fasting plasma insulin and C peptide were measured on an Immulite autoanalyzer (Diagnostic Products Corp, Los Angeles, CA). Wako (Richmond, VA) reagents were used for FFA determinations, except the FFAs obtained during the period of insulin infusion during the clamp, which were measured in triplicate by HPLC due to a requirement for greater sensitivity (23).
Animal studies
Twelve- to 14-month-old male C57BL/6J aP2-lipin-1b transgenic mice and nontransgenic littermates (6) were fed a diet containing 35% fat and 33% carbohydrate (Bioserve F3282; Bioserve, Frenchtown, NJ) for 6 wk. Epididymal fat pads were dissected and immediately frozen in liquid nitrogen for RNA isolation. Animal studies were performed under the approval of the University of California, Los Angeles institutional animal care and use committee.
RNA and DNA extraction
Total RNA from approximately 200 mg human or mouse adipose tissue was isolated with TRIzol reagent (Invitrogen, Carlsbad, CA), and human RNA was further purified on RNAEasy columns (QIAGEN, Valencia, CA). The quantity and integrity of the RNA were confirmed with the Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA).
Quantitative RT-PCR (qRT-PCR)
All primers and probes for human gene expression analysis were designed using Primer Express version 2.1 (Applied Biosystems, Roche, Branchburg, NJ). Sequences of primers and probes are shown in supplemental Table 1, which is published as supplemental data on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org. For the human studies, qRT-PCR reactions (24) were performed as one-step reactions in ABI PRISM 7900 (Applied Biosystems) using the following parameters: one cycle of 48 C for 30 min, then 95 C for 10 min, followed by 40 cycles at 95 C for 15 sec, and 60 C for 1 min. 18S rRNA was quantitated as a normalization control. For mouse gene expression analysis, qRT-PCR was performed with the iCycler (Bio-Rad Laboratories, Hercules, CA) using SYBR Green PCR reagents (QIAGEN). Gene expression was normalized to the housekeeping genes hypoxanthine phosphoribosyltransferase and β-2 microglobulin.
Statistical analysis
Population characteristics are represented as means ± sd. Gene expression and clinical data were correlated using regression analysis. ANOVA was used to test for differences in biopsy and blood parameters, with post hoc testing by mean equality contrast between different groups using the Tukey-Kramer HSD (α = 0.05). Type I error rate was set a priori at P < 0.05. Analysis was performed using JMP version 5.0 (SAS Institute Inc., Cary, NC).
Results
Lipin-1 mRNA expression levels are correlated with GDR, body fat, fat cell size, and FFAs in healthy young men
We studied 56 young healthy men (22.6 ± 3.2 yr; 26.4 ± 4.1 kg/m2) for a possible relationship between adipose tissue lipin-1 expression levels and parameters associated with adiposity and glucose homeostasis. The characteristics of our study population are shown in Table 1. Measurements reported here were performed while subjects were inpatients on a weight-maintaining diet. In studies described here, we quantitated expression levels of total lipin-1, as well as the individual lipin-1a and lipin-1b isoforms, which are produced from the LPIN1 gene through alternative mRNA splicing (25). Lipin-1a, lipin-1b, and total lipin-1 mRNA levels were highly intercorrelated, (R2 = 0.83–0.87), so for clarity of presentation, we refer to the total lipin-1 mRNA levels in the text, and include the experimental data for lipin-1a and lipin-1b isoforms in supplemental Tables 2–4, which are published as supplemental data on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org.
Table 1.
Characteristics of the study population
| Subject characteristics | Mean ± sd | Range |
|---|---|---|
| Age (yr) | 22.6 ± 3.2 | 18.0–29.0 |
| Height (cm) | 176.9 ± 5.8 | 163.0–189.5 |
| Weight (kg) | 82.5 ± 13.2 | 59.2–118.3 |
| BMI (kg/m2) | 26.4 ± 4.1 | 20.1–34.7 |
| WHR (au) | 0.87 ± 0.07 | 0.7–1.0 |
| Body fat (%) | 20.3 ± 6.5 | 8.4–32.3 |
| Visceral adipose tissue mass (kg) | 2.1 ± 1.3 | 0.5–5.7 |
| Fasting RQ (au) | 0.84 ± 0.04 | 0.74–0.95 |
| Steady-state RQ (au)a | 0.93 ± 0.04 | 0.85–1.03 |
| ΔRQ (au) | 0.09 ± 0.04 | 0.03–0.25 |
| VO2max (ml/kg·min) | 41.2 ± 7.2 | 23.5–59.2 |
| Fasting glucose (mg/dl) | 80.5 ± 5.4 | 66.0–90.0 |
| Steady-state glucose (mg/dl)a | 89.0 ± 4.5 | 78.0–102.0 |
| Fasting insulin (μU/ml) | 8.2 ± 4.6 | 2.6–22.4 |
| Steady-state insulin (μU/ml)a | 160.1 ± 35.7 | 103.9–251.3 |
| C peptide (ng/ml) | 1.93 ± 0.70 | 0.9–3.5 |
| GDR (mg/kg FFM·min) | 11.1 ± 4.1 | 4.0–24.5 |
| Fasting FFA (mmol) | 0.6 ± 0.2 | 0.07–0.84 |
| Steady-state FFA (mmol)a | 0.05 ± 0.02 | 0.02–0.16 |
| Fat cell size (μl) | 0.6 ± 0.2 | 0.22–0.95 |
All male subjects were chosen from the larger mixed gender study population. au, Arbitrary units; WHR, waist-to-hip ratio.
During insulin infusion (80 mIU/m2 BSA·min).
Adipose tissue lipin-1 mRNA expression levels were positively correlated with GDR during a euglycemic-hyperinsulinemic clamp (R2 = 0.22; P < 0.01) (Fig. 1A and Table 2). This correlation remained statistically significant even when analyzed without an apparent outlier (with a GDR of approximately 24 and a lipin mRNA of about 2.25), and when adjusted for adiposity (P = 0.03). A negative correlation was observed between lipin-1 expression levels and percent body fat, fat cell size, and fasting FFAs (Fig. 1, B–D). Lipin-1 was also negatively correlated with total adipose tissue mass (R2 = 0.20; P < 0.01), deep sc adipose tissue mass (R2 = 0.24; P < 0.01), sc adipose tissue mass (R2 = 0.20; P < 0.01), and visceral adipose tissue mass (R2 = 0.11; P < 0.05).
Figure 1.
Lipin-1 mRNA levels correlate with GDR, body fat, fat cell size, and FFAs. GDR was measured by a euglycemic-hyperinsulinemic clamp in the population of 56 healthy young men after an overnight fast. Body fat by DXA, fat cell size, and FFA were also measured in the fasting state. Lipin-1 mRNA expression levels were correlated with higher GDR (A), lower percent body fat (B), smaller fat cell size (C), and lower levels of fasting FFAs (D). mRNA expression data were normalized to 18S rRNA. Type I error rate was set a priori at P < 0.05.
Table 2.
Lipin-1 gene expression correlations with study population characteristics (R2)
| Subject characteristics | Lipin-1 (R2) | P value |
|---|---|---|
| BMI (kg/m2) | −0.17 | <0.01 |
| Body fat (%) | −0.19 | <0.01 |
| Total adipose tissue mass (kg) | −0.20 | <0.01 |
| Subcutaneous adipose tissue mass (kg) | −0.20 | <0.01 |
| Deep sc adipose tissue mass (kg) | −0.24 | <0.01 |
| Visceral adipose tissue mass (kg) | −0.11 | <0.05 |
| VO2max (ml/kg·min) | 0.16 | <0.01 |
| Fasting RQ (au) | NS | NS |
| Steady-state RQ (au)a | 0.16 | <0.01 |
| ΔRQ (au) | NS | NS |
| Fasting glucose (mg/dl) | NS | NS |
| Steady-state glucose (mg/dl)a | NS | NS |
| Fasting insulin (μU/ml) | NS | NS |
| Steady-state insulin (μU/ml)a | NS | NS |
| Fasting NEFA (μM) | −0.11 | <0.05 |
| Steady-state NEFA (μM)a | NS | NS |
| GDR (mg/kg FFM·min) | 0.22 | <0.01 |
| Adiponectin (μg/ml) | NS | NS |
| C peptide (ng/ml) | −0.10 | <0.05 |
| Triglycerides (mg/dl) | −0.10 | <0.05 |
| Leptin (mg/dl) | −0.11 | <0.05 |
| Fat cell size (μl) | −0.14 | <0.01 |
All male subjects were chosen from the larger mixed-gender study population. au, Arbitrary units; NEFA, non-esterified fatty acids; NS, nonsignificant; WHR, waist-to-hip ratio.
During insulin infusion (80 mIU/m2 BSA·min).
Relationship between lipin-1 expression and fasting and insulin-stimulated RQ
Previous indirect calorimetry studies of lipin-1 deficient mice implicated lipin-1 in metabolic switching between glucose and fatty acid substrates during fed and fasted states, as determined by measurements of RQ (26). To investigate whether a relationship exists between lipin-1 expression and substrate switching in humans, we measured RQ in fasted and insulin-stimulated states, before and during the euglycemic-hyperinsulinemic clamp. No significant correlation was observed between lipin-1 and fasting RQ (Fig. 2A). However, insulin-stimulated RQ at steady state during the clamp showed a positive correlation with lipin-1 expression (R2 = 0.16; P < 0.01) (Fig. 2B), consistent with increased insulin-stimulated glucose disposal. There was also a positive correlation between lipin-1 expression and maximum oxygen consumption during exercise (VO2max; R2 = 0.16; P < 0.01) (Table 2).
Figure 2.
Lipin-1 mRNA levels correlate with insulin-stimulated RQ. RQ was measured before and during a euglycemic-hyperinsulinemic clamp in the population of 56 healthy young men. A, Fasting RQ was not correlated with lipin-1 mRNA expression. B, However, insulin-stimulated RQ was positively correlated with lipin-1 mRNA. mRNA expression data were normalized to 18S rRNA. Type I error rate was set a priori at P < 0.05.
Correlation between lipin-1 and lipid metabolism gene expression levels in human and mouse adipose tissue
The results described previously indicate that there is a positive relationship between lipin-1 expression levels and a favorable metabolic profile, including improved insulin sensitivity, and reduced adipose tissue mass, adipocyte cell size, and FFA levels. Because lipin-1 functions as a transcriptional coactivator of PPARα in liver (10), we hypothesized that lipin-1 levels may exert similar effects on expression of PPARα and its target genes in adipose tissue, which could contribute to the changes in metabolic parameters described previously. To test this possibility, we investigated the potential relationship between lipin-1 expression levels and PPARα, as well as other genes with key roles in lipid metabolism, in adipose tissue from humans and from transgenic mice with enhanced lipin-1 expression.
In human adipose tissue, we observed a striking positive correlation between lipin-1 and PPARα mRNA levels (R2 = 0.74; P < 1 × 10−7) (Fig. 3A). The PPARα target gene medium-chain acyl-coenzyme A dehydrogenase (MCAD) also had a strong positive correlation with lipin-1 (R2 = 0.68; P < 1 × 10−7) (Fig. 3B). Lipin-1 also showed positive correlations with phosphoenol pyruvate carboxykinase (PCK1), fatty acid synthase (FAS), CD36, Cbl-associated protein (CAP), stearoyl coenzyme A desaturase-1 (SCD1), hormone-sensitive lipase (HSL), adipose tissue triglyceride lipase (ATGL), and perilipin (Table 3). As such, lipin-1 expression levels are correlated with genes involved in both fatty acid utilization/oxidation (PPARα, MCAD, CD36, ATGL, and HSL), and genes involved in lipid biosynthesis (SCD1 and FAS). Lipoprotein lipase (LPL) did not exhibit a significant correlation with lipin-1 expression, nor did several other genes investigated in these samples, such as adipose tissue inflammatory markers (CD68, MAC-2, monocyte chemoattractant protein-1, macrophage inhibitory protein-1 alpha; data not shown).
Figure 3.
Lipin-1 mRNA levels correlate with PPARα and MCAD expression levels. Expression levels for several lipid metabolism genes were measured by qRT-PCR (see Table 3). Of all lipid metabolism genes measured, PPARα (A) and MCAD (B) had the strongest correlation with lipin-1 levels. mRNA expression data were normalized to 18S rRNA. Type I error rate was set a priori at P < 0.05.
Table 3.
Lipin-1 gene expression correlations with adipose tissue gene expression (R2)
| Gene expression (mRNA) | Lipin-1 (R2) | P value |
|---|---|---|
| Lipid uptake | ||
| CD36 | 0.59 | 1 × 10−7 |
| LPL | 0.271 | 0.07 |
| Lipogenesis | ||
| FAS | 0.35 | 1.6 × 10−6 |
| PCK1 | 0.54 | 1 × 10−7 |
| SCD1 | 0.27 | 2.83 × 10−4 |
| Lipid oxidation | ||
| MCAD | 0.68 | 1 × 10−7 |
| PPARα | 0.74 | 1 × 10−7 |
| Lipolysis | ||
| Perilipin | 0.64 | 1 × 10−7 |
| CAP | 0.47 | 1 × 10−7 |
| ATGL | 0.61 | 1 × 10−7 |
| HSL | 0.68 | 1 × 10−7 |
All data are presented as R2. CD36, CD36 molecule (thrombospondin receptor).
The role of lipin-1 as a transcriptional coactivator raised the possibility that lipin-1 is a key driver in the expression of PPARα and other lipid metabolism genes that were positively correlated with lipin-1 expression in human adipose tissue. If true, we would predict that constitutively high lipin-1 expression levels in adipose tissue would lead to increased expression of PPARα and other genes. To address this possibility, we investigated whether the aP2-lipin-1 transgenic mouse, which exhibits a 4-fold constitutive increase in lipin-1 expression in adipose tissue (6) (Fig. 4A), also has elevated expression of genes showing a positive correlation with lipin-1 levels in human adipose tissue. Compared with nontransgenic littermates, the aP2-lipin-1 transgenic mice showed a significant increase in the expression of CD36, FAS, PPARα, MCAD, and HSL (P < 0.05) (Fig. 4A). As with human adipose tissue, we did not observe evidence for a relationship between lipin-1 and LPL, which had similar levels in wild-type and transgenic mice (data not shown). To exclude the possibility that the altered gene expression in the aP2-lipin-1 transgenic mice is related to the increased adiposity compared with nontransgenic animals (6), we also examined gene expression levels in another model of increased adiposity, the leptin-deficient ob/ob mouse. Unlike the lipin-1 transgenic animals, ob/ob mice had reduced mRNA levels compared with wild-type littermates for lipin-1, PPARα, MCAD, FAS, and HSL (Fig. 4B), indicating that the elevation of these genes does not occur simply as a consequence of increased adipose tissue mass. CD36 was increased in both aP2-lipin-1 transgenic and ob/ob mice, and may, therefore, be related to increased adiposity.
Figure 4.
Adipose tissue mRNA expression of lipid metabolism genes in aP2-lipin-1 transgenic and ob/ob mice. A, Real-time RT-PCR analysis of gene expression in epididymal adipose tissue of male aP2-lipin-1 transgenic (n = 4) and wild-type (wt) (n = 4) littermates. The average value for wild-type expression of each gene was set to one, and levels in transgenic mice were expressed relative to wild-type. *, P < 0.05. B, Real-Time RT-PCR analysis of gene expression in adipose tissue of ob/ob (n = 6) and wild-type (n = 5) mice. The value of wild type was set to one, and levels in ob/ob mice were expressed relative to wild type. *, P < 0.05. Expression data were normalized to hypoxanthine phosphoribosyl transferase and β2-microglobulin expression levels.
Discussion
Lipin-1 is a protein with roles in lipid biosynthesis and nuclear receptor coactivation, both of which may have important effects on lipid homeostasis and metabolism. Previous studies in vitro and in vivo have linked lipin-1 to the maintenance of metabolic homeostasis. The lipin-1 deficient fatty liver dystrophy mouse and adipose tissue-specific transgenic models have shown that lipin-1 is a determinant of adiposity, energy expenditure, and insulin sensitivity (4,6). Recent studies in human adipose tissue have revealed correlations between lipin-1 levels and improved glucose homeostasis but have involved dyslipidemic or obese, glucose intolerant subjects (11,12). This study represents the first to investigate whether lipin-1 expression levels are associated with glucose homeostasis in healthy young men, and the first to use direct measurements of insulin sensitivity by a euglycemic-hyperinsulinemic clamp, as well as characterization of energy substrate utilization by indirect calorimetry. In addition, we examined the relationship between human lipin-1 gene expression and PPARα, which recently was shown to be regulated by lipin-1 in mouse liver (10).
Lipin-1 was positively correlated with GDR during a euglycemic-hyperinsulinemic clamp, providing the most direct evidence so far that lipin-1 mRNA levels in adipose tissue are positively correlated with insulin sensitivity. Furthermore, these results demonstrate that this relationship holds true even in healthy young subjects. Together with previous studies, these results establish that enhanced lipin-1 expression levels in adipose tissue are associated with insulin sensitivity over a broad spectrum of BMI, ranging from lean to obese, and healthy to glucose-intolerant individuals (11,12). There are several possible mechanisms for the observed relationship between lipin-1 expression and insulin sensitivity. Previous studies revealed that lipin-1 expression levels in adipose tissue are correlated with both basal and insulin-stimulated glucose uptake in human adipocytes (13), and inversely correlated with lipid accumulation in skeletal muscle (12). Here, we detected a positive correlation between lipin-1 mRNA and insulin-stimulated RQ, suggesting that lipin-1 is associated with the ability to switch substrate oxidation from fat to carbohydrate upon insulin stimulation.
Consistent with its positive correlation with insulin sensitivity, lipin-1 also showed a negative correlation with measures of adiposity, including total, visceral, and sc adipose tissue mass, as well as fat cell size and BMI. Previous studies have also noted an inverse correlation between lipin-1 and BMI (11,12,27). Given the known role of lipin-1 in triglyceride biosynthesis in adipose tissue, these results may at first appear counterintuitive. However, they are in line with the demonstration that adipose tissue lipogenic capacity and lipogenic enzyme gene expression are decreased in human obesity (28,29). In addition, emerging information about the function of lipin-1 indicates that its molecular role extends beyond its PAP1 enzyme function to include gene regulatory activity (10). Indeed, the current study provides in vivo evidence for a correlation between lipin-1 mRNA levels and the expression in human adipose tissue of PPARα and other genes that have been implicated as regulatory targets (discussed below). Thus, the mechanisms by which lipin-1 mRNA levels influence adiposity and insulin sensitivity may be clarified as the biochemical and molecular processes that regulate the dual actions of lipin-1 are elucidated.
We also detected a strong positive correlation between expression levels of lipin-1 and genes with key roles in fatty acid oxidation—PPARα and MCAD. This relationship held true in transgenic mouse adipose tissue, in which enhanced lipin-1 expression driven by a heterologous promoter led to enhanced PPARα and MCAD expression. In the aP2-lipin-1 transgenic mice, the only difference from wild-type mice at the genomic level is the lipin-1 transgene, which confers a 3- to 4-fold increase in lipin-1 mRNA levels in adipose tissue. Thus, changes in gene expression in the adipose tissue of this model are either directly or indirectly related to the increased lipin-1 expression. Although these correlations between lipin-1 and PPARα expression do not establish mechanism, it is relevant that in mouse liver, lipin-1 expression induces the expression of PPARα and also acts as a coactivator for PPARα, resulting in the activation of fatty acid oxidation genes (10). Thus, our results raise the possibility that lipin-1 may play an analogous role in adipose tissue, acting as a regulator of PPARα expression and/or coactivator activity. In that case, enhanced lipin-1 expression could lead to increased fatty acid oxidation in adipose tissue, and improved metabolic homeostasis. In addition, there were strong positive correlations between lipin-1 and lipid storage genes (PCK1, SCD1 and FAS) and lipolysis genes (perilipin, ATGL, HSL, and CAP). It remains to be determined whether lipin-1 can directly modulate expression of these genes via its transcriptional coactivator function.
The one discrepancy that we have noted between human and mouse is that in human adipose tissue, lipin-1 levels were inversely correlated with adiposity, whereas in the aP2-lipin-1 transgenic mouse, we previously showed a positive correlation (6). We do not know the basis for this difference, but it may be related to the fact that the increased lipin-1 gene expression in aP2-lipin-1 transgenic mice is constitutively up-regulated via heterologous regulatory elements and, therefore, not responsive to molecular or physiological signals that normally modulate lipin-1 levels. Because lipin-1 is a required PAP1 enzyme for triacylglycerol synthesis in adipose tissue, forced expression may be responsible for the enhanced triglyceride accumulation in adipose tissue of lipin-1 transgenic mice, whereas in normal human subjects, lipin-1 gene expression in fat cells may be attenuated upon reaching a normal, mature fat cell size. Support for this possibility comes from studies demonstrating that in cases in which human adipose tissue is not at its normal triglyceride storage capacity, as in lipodystrophy, lipin-1 expression levels are positively correlated with fat mass (30).
In summary, we provide evidence that lipin-1 expression levels in adipose tissue are correlated with adiposity, insulin sensitivity, and oxidative, lipogenic, and lipolytic gene expression in adipose tissue from healthy young human subjects. Correlations between lipin-1 gene expression levels and that of several other lipid metabolism genes were also observed in adipose tissue from aP2-lipin-1 transgenic mice, suggesting that expression of these genes may be driven either directly or indirectly by lipin-1. Notably, the expression levels of PPARα, a target of lipin-1 transcriptional coactivator activity, were strongly positively correlated with lipin-1 levels in human and transgenic mouse adipose tissue. Induction of PPARα and its target genes involved in fatty acid oxidation may be a mechanism by which lipin-1 levels influence insulin sensitivity.
Supplementary Material
Footnotes
First Published Online October 9, 2007.
Abbreviations: ATGL, Adipose tissue triglyceride lipase; BMI, body mass index; CAP, Cbl-associated protein; DXA, dual-energy x-ray absorptiometry; FAS, fatty acid synthase; FFA, free fatty acid; FFM, fat free mass; GDR, glucose disposal rate; HSL, hormone-sensitive lipase; Lpin1, lipin-1; LPL, lipoprotein lipase; MCAD, medium-chain acyl-coenzyme A dehydrogenase; PAP1, phosphatidate phosphatase type-1; PCK1, phosphoenol pyruvate carboxykinase; PPAR, peroxisome proliferator-activated receptor; qRT-PCR, quantitative RT-PCR, respiratory quotient; SCD1, stearoyl coenzyme A desaturase-1; VO2max, maximal aerobic capacity.
This work was supported by National Institutes of Health Grants HL-28481 (to K.R.), DK-72476 (to S.R.S.), and Genomic Analysis Training Grant S-T32-HG002536 (to J.D.), and the United States Department of Agriculture Grant 58-6435-5-017 (to S.R.S.).
Disclosure Statement: The authors have nothing to disclose.
References
- Geiss LS, Pan L, Cadwell B, Gregg EW, Benjamin SM, Engelgau MM 2006 Changes in incidence of diabetes in U.S. adults, 1997–2003. Am J Prev Med 30:371–377 [DOI] [PubMed] [Google Scholar]
- York DA, Rossner S, Caterson I, Chen CM, James WP, Kumanyika S, Martorell R, Vorster HH 2004 Prevention Conference VII: Obesity, a worldwide epidemic related to heart disease and stroke: group I: worldwide demographics of obesity. Circulation 110:e463–e470 [DOI] [PubMed] [Google Scholar]
- Peterfy M, Phan J, Xu P, Reue K 2001 Lipodystrophy in the fld mouse results from mutation of a new gene encoding a nuclear protein, lipin. Nat Genet 27:121–124 [DOI] [PubMed] [Google Scholar]
- Reue K, Xu P, Wang XP, Slavin BG 2000 Adipose tissue deficiency, glucose intolerance, and increased atherosclerosis result from mutation in the mouse fatty liver dystrophy (fld) gene. J Lipid Res 41:1067–1076 [PubMed] [Google Scholar]
- Phan J, Peterfy M, Reue K 2004 Lipin expression preceding peroxisome proliferator-activated receptor-γ is critical for adipogenesis in vivo and in vitro. J Biol Chem 279:29558–29564 [DOI] [PubMed] [Google Scholar]
- Phan J, Reue K 2005 Lipin, a lipodystrophy and obesity gene. Cell Metab 1:73–83 [DOI] [PubMed] [Google Scholar]
- Han GS, Wu WI, Carman GM 2006 The Saccharomyces cerevisiae Lipin homolog is a Mg2+-dependent phosphatidate phosphatase enzyme. J Biol Chem 281:9210–9218 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Donkor J, Sariahmetoglu M, Dewald J, Brindley DN, Reue K 2007 Three mammalian lipins act as phosphatidate phosphatases with distinct tissue expression patterns. J Biol Chem 282:3450–3457 [DOI] [PubMed] [Google Scholar]
- Harris TE, Huffman TA, Chi A, Shabanowitz J, Hunt DF, Kumar A, Lawrence Jr JC 2007 Insulin controls subcellular localization and multisite phosphorylation of the phosphatidic acid phosphatase, lipin 1. J Biol Chem 282:277–286 [DOI] [PubMed] [Google Scholar]
- Finck BN, Gropler MC, Chen Z, Leone TC, Croce MA, Harris TE, Lawrence Jr JC, Kelly DP 2006 Lipin 1 is an inducible amplifier of the hepatic PGC-1α/PPARα regulatory pathway. Cell Metab 4:199–210 [DOI] [PubMed] [Google Scholar]
- Suviolahti E, Reue K, Cantor RM, Phan J, Gentile M, Naukkarinen J, Soro-Paavonen A, Oksanen L, Kaprio J, Rissanen A, Salomaa V, Kontula K, Taskinen MR, Pajukanta P, Peltonen L 2006 Cross-species analyses implicate Lipin 1 involvement in human glucose metabolism. Hum Mol Genet 15:377–386 [DOI] [PubMed] [Google Scholar]
- Yao-Borengasser A, Rasouli N, Varma V, Miles LM, Phanavanh B, Starks TN, Phan J, Spencer 3rd HJ, McGehee Jr RE, Reue K, Kern PA 2006 Lipin expression is attenuated in adipose tissue of insulin-resistant human subjects and increases with peroxisome proliferator-activated receptor γ activation. Diabetes 55:2811–2818 [DOI] [PubMed] [Google Scholar]
- van Harmelen V, Ryden M, Sjolin E, Hoffstedt J 2007 A role of lipin in human obesity and insulin resistance: relation to adipocyte glucose transport and GLUT4 expression. J Lipid Res 48:201–206 [DOI] [PubMed] [Google Scholar]
- Sparks LM, Xie H, Koza RA, Mynatt R, Hulver MW, Bray GA, Smith SR 2005 A high-fat diet coordinately downregulates genes required for mitochondrial oxidative phosphorylation in skeletal muscle. Diabetes 54:1926–1933 [DOI] [PubMed] [Google Scholar]
- Sparks LM, Xie H, Koza RA, Mynatt R, Bray GA, Smith SR 2006 High-fat/low-carbohydrate diets regulate glucose metabolism via a long-term transcriptional loop. Metabolism 55:1457–1463 [DOI] [PubMed] [Google Scholar]
- Ukropcova B, McNeil M, Sereda O, de Jonge L, Xie H, Bray GA, Smith SR 2005 Dynamic changes in fat oxidation in human primary myocytes mirror metabolic characteristics of the donor. J Clin Invest 115:1934–1941 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ukropcova B, Sereda O, de Jonge L, Bogacka I, Nguyen T, Xie H, Bray GA, Smith SR 2007 Family history of diabetes links impaired substrate switching and reduced mitochondrial content in skeletal muscle. Diabetes 56:720–727 [DOI] [PubMed] [Google Scholar]
- Heilbronn LK, Rood J, Janderova L, Albu JB, Kelley DE, Ravussin E, Smith SR 2004 Relationship between serum resistin concentrations and insulin resistance in nonobese, obese, and obese diabetic subjects. J Clin Endocrinol Metab 89:1844–1848 [DOI] [PubMed] [Google Scholar]
- Carter H, Jones AM, Barstow TJ, Burnley M, Williams CA, Doust JH 2000 Oxygen uptake kinetics in treadmill running and cycle ergometry: a comparison. J Appl Physiol 89:899–907 [DOI] [PubMed] [Google Scholar]
- Smith SR, Lovejoy JC, Greenway F, Ryan D, deJonge L, de la Bretonne J, Volafova J, Bray GA 2001 Contributions of total body fat, abdominal subcutaneous adipose tissue compartments, and visceral adipose tissue to the metabolic complications of obesity. Metabolism 50:425–435 [DOI] [PubMed] [Google Scholar]
- Livesey G, Elia M 1985 Food energy values of artificial feeds for man. Clin Nutr 4:99–111 [DOI] [PubMed] [Google Scholar]
- Smith SR, Xie H, Baghian S, Needham A, McNeil M, Bogacka I, Bray GA 2006 Pioglitazone changes the distribution of adipocyte size in type 2 diabetics. Adipocytes 2:11–22 [Google Scholar]
- Miles JM, Ellman MG, McClean KL, Jensen MD 1987 Validation of a new method for determination of free fatty acid turnover. Am J Physiol 252(3 Pt 1):E431–E438 [DOI] [PubMed] [Google Scholar]
- Bustin SA 2000 Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. J Mol Endocrinol 25:169–193 [DOI] [PubMed] [Google Scholar]
- Peterfy M, Phan J, Reue K 2005 Alternatively spliced lipin isoforms exhibit distinct expression pattern, subcellular localization, and role in adipogenesis. J Biol Chem 280:32883–32889 [DOI] [PubMed] [Google Scholar]
- Xu J, Lee WN, Phan J, Saad MF, Reue K, Kurland IJ 2006 Lipin deficiency impairs diurnal metabolic fuel switching. Diabetes 55:3429–3438 [DOI] [PubMed] [Google Scholar]
- Croce MA, Eagon JC, Lariviere LL, Korenblat KM, Klein S, Finck BN 2007 Hepatic lipin-1b expression is diminished in insulin-resistant obese subjects and is reactivated by marked weight loss. Diabetes 56:2395–2399 [DOI] [PubMed] [Google Scholar]
- Ranganathan G, Unal R, Pokrovskaya I, Yao-Borengasser A, Phanavanh B, Lecka-Czernik B, Rasouli N, Kern PA 2006 The lipogenic enzymes DGAT1, FAS, and LPL in adipose tissue: effects of obesity, insulin resistance, and TZD treatment. J Lipid Res 47:2444–2450 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Diraison F, Dusserre E, Vidal H, Sothier M, Beylot M 2002 Increased hepatic lipogenesis but decreased expression of lipogenic gene in adipose tissue in human obesity. Am J Physiol Endocrinol Metab 282:E46–E51 [DOI] [PubMed] [Google Scholar]
- Lindegaard B, Larsen LF, Hansen AB, Gerstoft J, Pedersen BK, Reue K 2007 Adipose tissue lipin expression levels distinguish HIV patients with and without lipodystrophy. Int J Obes (Lond) 31:449–456 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.