Human Molecular Genetics Advance Access originally published online on December 15, 2005
Human Molecular Genetics 2006 15(3):377-386; doi:10.1093/hmg/ddi448
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cross-species analyses implicate Lipin 1 involvement in human glucose metabolism
1Department of Molecular Medicine, National Public Health Institute and 2Department of Medical Genetics, University of Helsinki, Biomedicum Helsinki, Helsinki, Finland, 3Department of Human Genetics and 4Department of Pediatrics, David Geffen School of Medicine at UCLA, Los Angeles, USA, 5Biomedicum Bioinformatics Unit, 6Department of Medicine and 7Finnish Twin Cohort Study, Department of Public Health, University of Helsinki, Helsinki, Finland, 8Department of Mental Health and Alcohol Research and 9Department of Epidemiology and Health Promotion, National Public Health Institute, Helsinki, Finland, 10Obesity Research Unit, Helsinki University Central Hospital, Helsinki, Finland and 11The Broad Institute, MIT, Boston, MA, USA
* To whom correspondence should be addressed at: Biomedicum Helsinki, PO Box 104, 00251 Helsinki, Finland. Tel: +358 947448393; Fax: +358 947448480; Email: leena.peltonen{at}ktl.fi
Received August 5, 2005; Accepted December 3, 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Recent studies in the mouse have demonstrated that variations in lipin expression levels in adipose tissue have marked effects on adipose tissue mass and insulin sensitivity. In the mouse, lipin deficiency prevents normal adipose tissue development, resulting in lipodystrophy and insulin resistance, whereas excess lipin levels promote fat accumulation and insulin sensitivity. Here, we investigated the effects of genetic variation in lipin levels on glucose homeostasis across species by analyzing lipin transcript levels in human and mouse adipose tissues. A strong negative correlation was observed between lipin mRNA levels and fasting glucose and insulin levels, as well as an indicator of insulin resistance (HOMA-IR), in both mice and humans. We subsequently analyzed the allelic diversity of the LPIN1 gene in dyslipidemic Finnish families, as well as in a case–control sample of obese (n=477) and lean (n=821) individuals. Alleles were defined by genotyping seven single nucleotide polymorphisms (SNPs) of the critical DNA region over the LPIN1 gene. Intragenic SNPs and corresponding allelic haplotypes exhibited associations with serum insulin levels and body mass index (P=0.002–0.04). Both the expression levels in adipose tissue across species and genetic data in human study samples highlight the importance of lipin in glucose homeostasis and imply that allelic variants of this gene have significance in human metabolic traits.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Obesity contributes to a number of serious health problems in western societies, as it increases the risk for several common diseases including type 2 diabetes, coronary heart disease, hypertension, osteoarthritis and certain forms of cancer (reviewed in 1
,2). In particular, the accumulation of visceral fat (abdominal obesity) is often accompanied by insulin resistance, elevated serum triglycerides (TGs), reduced high-density lipoprotein cholesterol (HDL-C) and hypertension, all of which are components of the metabolic syndrome.
Adipose tissue plays a critical role in the regulation of lipid and glucose metabolism. Both in humans and in animal models, it has been shown that either increased adipose tissue mass as seen in obesity or abnormally low amounts of adipose tissue as seen in lipodystrophy, lead to metabolic dysregulation and insulin resistance (reviewed in 3–5). One mechanism by which impaired adipose tissue function influences metabolism is through secretion of adipokines, which have effects on energy metabolism and insulin action (6). Another mechanism is by influencing fat storage in non-adipose tissues, such as skeletal muscle or pancreatic ß cells, where it can be cytotoxic (7). Thus, factors that influence adipose tissue mass and function exert important effects on metabolic homeostasis.
Rodent data provide evidence that one factor involved in the development and function of adipose tissue is lipin. Lipin is a novel protein recently identified through positional cloning of the mutated gene (Lpin1, Entrez GeneID 14245) in a mouse model of lipodystrophy known as fatty liver dystrophy (fld) (8). Lipin deficiency in the fld mouse results in an 80% reduction in adipose tissue mass, mild hyperglycemia and insulin resistance (9). Similar to human lipodystrophic subjects, the fld mouse also exhibits a fatty liver and hypertriglyceridemia, but unlike human lipodystrophic subjects, this is transient, occurring only during the neonatal period (10). We have recently elucidated the basis for lipodystrophy in the fld mouse, which is related to a requirement for lipin early during the process of adipocyte differentiation (11). The adipose tissue in lipin-deficient mice is therefore composed of immature adipocytes incapable of normal fat storage and adipokine secretion.
Lipin is also expressed in mature adipocytes. As revealed by both in vitro studies and analysis of transgenic mice, elevated lipin expression leads to increased lipogenesis and TG accumulation in fat cells (12). Furthermore, in contrast to the insulin resistance observed in lipin-deficient mice, adipose-specific lipin transgenic mice exhibit increased insulin sensitivity compared with wild-type animals (12). Although the mechanism for the effect of increased lipin expression on insulin sensitivity is not known, possibilities include protection of non-adipose tissues from lipid accumulation due to preferential lipid storage in adipose tissue or increased levels of insulin sensitizing adipokine production in lipin transgenic cells.
On the basis of the observations in the mouse, we hypothesized that genetic variations in lipin expression levels and in the human LPIN1 gene locus (MIM 605518 [OMIM] , Entrez Gene ID 23175) might contribute to the regulation of glucose and lipid metabolism. Here we report a correlation between lipin mRNA expression levels in adipose tissue and serum glucose and insulin levels in both man and mouse. To provide additional evidence for the role of the lipin gene in human dyslipidemias or obesity, we studied the allelic diversity of LPIN1 in Finnish study samples ascertained for hyperlipidemias or obesity, and phenotyped for a wide spectrum of quantitative traits relevant to glucose and lipid metabolism. We found that specific alleles of LPIN1 were associated with serum insulin levels, implicating genetic variations in LPIN1 as a player in glucose homeostasis.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Lpin1 mRNA levels and glucose metabolism traits in mice
We first studied whether a relationship exists between lipin mRNA expression levels and glucose metabolism traits inanimals carrying a wild-type Lpin1 gene. We used twoapproaches to generate groups of mice having individual variation in glucose metabolism parameters and/or lipin expression levels. In the first approach, we introduced individual variation by producing a mixed genetic background in F2 mice derived from a cross between two inbred strains, C57BL/6J and BALB/cByJ. In the second approach, we fed a high fat diet to C57BL/6J mice to promote the development of obesity, hyperglycemia and hyperinsulinemia (13
). When mice of mixed genetic background were fed a chow diet, we found that lipin mRNA levels in gonadal adipose tissue varied up to 2-fold among individual animals, with levels tending to be higher in females than in males (Fig. 1). Body weight varied by 20–30% among individuals of both sexes, but there was no correlation between lipin levels and body weight for either sex. However, we did observe a negative correlation between lipin mRNA and fasting glucose levels in both sexes (Spearman correlation coefficient for males and females combined, r=–0.79, P<0.0001), with higher lipin expression levels associated with lower glucose levels (Fig. 1B). Negative correlations were also observed between lipin and insulin levels (r=–0.45, P<0.04) and with the HOMA-IR index of insulin resistance (r=–0.66, P<0.002). As seen in Figure 1C and D, the range of insulin and HOMA values in females was much smaller than in males and correlations were strongest in male mice, but reached statistical significance only when both sexes were combined.
|
Despite their genetic identity, C57BL/6J mice fed a high fat diet for 6 weeks exhibit inter-individual variation in diet response as measured by weight gain, glucose and insulin levels (14
). Similar to the observations in the chow-fed mixed background mice, the lipin mRNA levels among animals varied about 2-fold (Fig. 2). All mice gained weight on the high fat diet, but varied in their post-diet body weight up to 25% (Fig. 2A). In contrast to results for chow-fed mice, lipin levels in the fat-fed animals were negatively correlated tobody weight (r=–0.67, P<0.002). Interestingly, the observed correlation was almost entirely due to male mice(r=–0.69, P=0.03), with no significant correlation in females. As seen in the chow-fed mice with a mixed genetic background, there was a strong negative correlation between lipin mRNA levels and fasting glucose values in both sexes (Fig. 2; r=–0.82; P<0.0001). Insulin levels and HOMA-IR were also negatively correlated with lipin expression (r=–0.64, P<0.003 for insulin; r=–0.69, P<0.0008 for HOMA-IR). As seen in the chow-fed mice, the ranges of values were larger and correlations between lipin and insulin/HOMA were stronger in male animals, but significant only in the combined males and females (Fig. 2C and D).
|
LPIN1 transcript levels in human fat biopsies
To determine whether similar relationships between lipin expression and metabolic parameters occur in humans, we quantified lipin mRNA levels in 19 human fat biopsies (Table 1) using DNA microarray analysis. There was no correlation between LPIN1 expression levels and body mass index (BMI) (Fig. 3). However, similar to the findings in mice, strong negative correlations were observed between human lipin mRNA levels and glucose (r=–0.81, P=0.001), insulin (r=–0.74, P=0.001) and TG levels (r=–0.64, P=0.003). Lipin mRNA levels were also strongly negatively correlated with the HOMA-IR index of insulin resistance (Spearman correlation coefficient of –0.82; P=0.001). Thus, the negative correlation initially observed in the mouse between lipin expression levels in fat tissue and glucose, insulin and HOMA-IR, was also observed in humans.
|
|
Allelic diversity of the LPIN1 gene: LD and haplotype blocks
To assess the genetic variation in the LPIN1 gene, we characterized single nucleotide polymorphisms (SNPs) and their haplotype structure. We genotyped seven SNPs and estimated the linkage disequilibrium (LD) among them (Fig. 4) using the founders in the familial combined hyperlipidemia (FCHL) and low HDL-C families and lean control individuals as a study sample of independent individuals. SNPs 2–4 located within 28 kb between exon 1 and exon 6 of the LPIN1 gene were in considerable LD (D'>0.9) (Fig. 4). Altogether, the seven SNPs analyzed formed eight different haplotypes with frequencies greater than 5% in these Finnish study samples. The observed LD pattern is in agreement with the current data obtained from the CEPH families in the HapMap project (http://www.hapmap.org).
|
LPIN1 alleles in dyslipidemic families
We next assessed whether specific alleles of the LPIN1 gene locus are associated with human metabolic traits in the 92 carefully phenotyped dyspidemic families. We tested the seven SNPs spanning an 80 kb DNA region and their corresponding haplotypes for association with a set of quantitative traits derived from the phenotype of the fld mouse (9
,10). These included BMI, fasting TG, glucose and insulin levels, as well as lipoprotein lipase (LPL) and hepatic lipase (HL) activities in all the family members (n=1109). Females (n=571) and males (n=538) were also analyzed separately because these traits showed significant sex differences in our study sample (P<0.05). As shown in Table 2, an association with serum insulin levels was observed with SNP2 (P=0.008). The sex-specific analyses showed that primarily males contributed to the association (P=0.01). In females, HL activity was associated with SNP3 (P=0.02). We also performed haplotype analysis of seven SNPs and observed evidence for association with insulin levels in males (P=0.03).
|
To characterize more carefully the alleles associated with insulin levels and HL activity, we performed analyses using the family based association test (FBAT) program and monitored the transmission of different alleles and allelic haplotypes. The T-allele of SNP2 (with a frequency of 40%) was preferentially transmitted to those with elevated insulin levels and the G-allele of SNP3 (with a frequency of 55%) to those with elevated HL activity. The most common allelic haplotype 1 (A-T-C-T-G-C-C) with a frequency of 20% and the haplotype 10 (A-T-C-T-G-T-C) with a frequency of 4% were preferentially transmitted to individuals with elevated insulin levels. In contrast, the allelic haplotype 8 (A-T-C-T-A-C-T) with a frequency of 5%, was undertransmitted to those with higher insulin values. The allele frequencies of all the SNPs and SNP haplotypes are shown in Table 3.
|
To obtain empiric P-values for the allelic haplotype analysis with SNPs in considerable LD, we permuted the analysis 100 000 times using the haplotype permutation option in the FBAT program. The association with insulin levels remained significant in males for the overall seven SNP haplotype (P=0.04) as well as for haplotype 1 (A-T-C-T-G-C-C, P=0.05), haplotype 8 (A-T-C-T-A-C-T, P=0.002) and haplotype 10 (A-T-C-T-G-T-C, P=0.008).
Allelic diversity of LPIN1 in obese and lean individuals
After obtaining evidence that LPIN1 may be critical for the insulin levels and abnormal lipid traits, we wanted to study whether the variation in this gene is associated with obesity in the case–control sample, collected from the same population as the dyslipidemic families. We examined the association between LPIN1 genotypes and obesity in lean (BMI 20–25 kg/m2) and obese (BMI>30 kg/m2) Finnish subjects (Table 4) by testing the seven LPIN1 SNPs for differences in allele frequency between obese (n=493) and lean individuals (n=821). A difference in allele frequency was observed for SNP 6 (C/T) (0.11 versus 0.08 for the obese and lean, respectively, P=0.02; odds ratio of 1.4; 95% confidence interval 1.04–1.86). We also analyzed the LPIN1 SNPs for differences in mean BMI levels between different genotype groups. SNP1, SNP5 and SNP6 were associated with BMI in lean males (P-values of 0.01, 0.008 and 0.03, respectively) and SNP2,SNP3, SNP4 and SNP5 in obese males (P-values of 0.04, 0.008, 0.008 and 0.002, respectively). As no association was observed in lean or obese females, the association of LPIN1 alleles with BMI appears to be sex specific.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Although a number of inbred animal models with monogenic gene defects have provided novel insights into metabolic and regulatory mechanisms, establishing the functional role of these genes for human disease processes has remained a challenge. Here we provide evidence that a gene identified in a spontaneous mutant mouse strain with an aberrant metabolic phenotype also has relevance for human dyslipidemias and obesity, possibly by affecting insulin sensitivity in peripheral tissues such as fat.
Proper function of the Lpin1 gene is crucial for normal adipose tissue differentiation as well as for maintaining glucose and lipid homeostasis in the mouse (9,11,12). Our data give further evidence for a critical role of Lpin1 in glucose homeostasis in the mouse by demonstrating a negative correlation between lipin mRNA expression levels in adipose tissue and glucose levels, insulin levels and HOMA-IR. This was true in both inbred and mixed background mice, indicating a robust relationship between Lpin1 and glucose and insulin levels. In chow-fed mice, the correlation between lipin expression levels and glucose and insulin occurred independent of body weight, but in animals fed a high fat diet, there was a negative correlation between body weight and lipin expression. These results extend our previous findings showing that lipin deficiency in the fld mouse is associated with insulin resistance, whereas lipin over-expression in transgenic mice has the opposite effect (9,12). Together, these results demonstrate that an inverse correlation between lipin expression levels and glucose metabolism parameters occurs over a large range of lipin levels and strongly implicate Lpin1 as a determinant of glucose homeostasis in the mouse.
Stimulated by the mouse data, we extended our study of the LPIN1 gene to humans. Parallel to the findings in the mouse, the TG, glucose and insulin levels, as well as the HOMA-IR index, were negatively correlated with lipin mRNA expression levels in human adipose tissue. Overall, the striking consistency between the mouse and human data showing a robust negative correlation between lipin expression levels and glucose levels, and a less significant, but consistent, negative correlation with insulin levels suggests an important role for adipose tissue lipin levels as a modulator of insulin sensitivity in mammalian species in general.
Despite accumulating evidence from mouse studies that lipin plays an important role in adipose tissue development and glucose homeostasis, only one previous study has investigated whether variants of the human LPIN1 gene are associated with any human disease process. Cao and Hegele (15) studied the LPIN1 gene as a candidate gene for human lipodystrophy, but failed to identify any disease causing mutations. However, it has been shown that the region of chromosome 2p21 harboring LPIN1 is linked to variations in fat mass and leptin levels in Mexican-American, African-American and French populations (16–18). LPIN1 is located within a 25 Mb interval from the linkage peaks. Here we studied allelic distribution of the LPIN1 gene in dyslipidemic families to determine whether specific alleles of this gene were associated with measures of glucose and lipid metabolism. FCHL patients share some phenotypic features with the fld mouse, most notably insulin resistance. We detected significant associations between insulin levels and SNP2, as well as with SNP haplotypes in the combined study sample of 53 FCHL and 39 low HDL-C families. The association was strongest in males. Interestingly, neither FCHL nor low HDL-C families alone produced the associations observed, but both contributed to it (the most significant association for insulin levels in FCHL families for SNP2, P-value of 0.003; and in HDL-C families for the seven SNP haplotype, P-value of 0.04). This suggests that the relevance of the LPIN1 gene may not be restricted to a specific type of dyslipidemia, but instead may play a more general role in molecular pathways affecting glucose homeostasis, which are potentially shared by multiple clinical phenotypes, including FCHL and low HDL-C.
The hypothesis of a more general involvement of LPIN1 in glucose metabolism is further supported by our data from the case–control sample for obesity, showing a difference in allele frequencies of the LPIN1 gene as well as in BMI between different genotype groups in males. These results, taken together with the data obtained from hyperlipidemic families showing a stronger association with insulin levels in males, suggest that sex might influence the effect of the LPIN1 gene in glucose homeostasis. Whether this effect is tissue- or cell type-specific remains to be established in studies with a wider selection of tissue samples. Our results in the mouse also reveal a stronger effect in males, suggesting that this animal model is pertinent for the further elucidation of the role of lipin in glucose metabolism.
Our data provide the first evidence for the involvement of the LPIN1 gene in the regulation of glucose and insulin levels in humans. In particular, the strong negative correlation between lipin expression in adipose tissue and glucose and insulin levels suggests that higher lipin levels in normal or hyperlipidemic individuals may confer protection against the development of insulin resistance. Previous studies demonstrate that lipin phosphorylation occurs in response to insulin via activation of the mammalian target of rapamycin (mTOR) pathway (19). Given that mTOR plays a central role in nutrient-sensing to maintain a proper balance between amino acid availability, protein synthesis and cell growth (20), it is possible that lipin may act as a downstream effector of mTOR signaling to influence the response to insulin. In addition, the actions of lipin may not be limited to its effects on peripheral tissue insulin sensitivity as Lpin1 expression has also been detected in pancreatic beta cells (21). Thus, it is also possible that lipin levels may modulate glucose homeostasis by directly influencing insulin secretion. Taken together our data should encourage further studies on the role of lipin genes in the molecular pathogenesis of human metabolic diseases and their trait components.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Mouse studies
Inbred C57BL/6J and BALB/cByJ mice were obtained from The Jackson Laboratory (Bar Harbor, ME). All mice were maintained on a 12 h light/dark cycle and housed individually beginning at 4 weeks of age. Studies performed here used either C57BL/6J mice or a mixed genetic background produced in an F2 cross between C57BL/6J and BALB/cByJ. A broad-sense heritability estimate for this cross was calculated as the coefficient of genetic determination (g2) for lipin mRNA levels in male chow-fed C57BL/6J and BALB/cByJ mice described in Turner et al. (22
). The resulting coefficient of genetic determination was 0.29 (P<0.05; n=6 for each strain).
For studies on the mixed genetic background, mice were fed standard mouse chow containing 
4% fat, and blood and tissues harvested at 4 months of age. For studies of C57BL/6J mice on a high fat diet, 6-month-old animals were fed a diet containing 35% fat and 33% carbohydrate (Diet F3282, Bio-Serve, Frenchtown, NJ, USA) for 6 weeks. For glucose and insulin determinations, mice were fasted 12 h before blood collection. Glucose levels were determined in whole blood using a One Touch Ultra Blood Glucose Monitor (Lifescan, Milpitas, CA, USA). Insulin levels were determined in 5 µl plasma by Ultrasensitive Mouse Insulin EIA (ALPCO Diagnostics, Windham, MA, USA). Adipose tissue was snap-frozen for RNA isolation.
Lipin 1 expression analysis in mouse and human adipose tissue
Mouse adipose tissue RNA was isolated from gonadal fat padsusing Trizol (Invitrogen, Carlsbad, CA, USA), and cDNA synthesized using oligo dT primers. Real-time PCR reactions were performed on the iCycler iQ Real-Time Detection System (BioRad) using SYBR Green PCR QuantiTect reagent kit (Qiagen, Valencia, CA, USA). Each assay included (in triplicate): a standard curve of four serial dilution points of control cDNA (ranging from 100 ng to 100 pg), a no-template control and 25–50 ng of each sample cDNA. Primers for Lpin1 and for endogenous control genes HPRT (hypoxanthine phosphoribosyltransferase) and TBP (TATA box binding protein) were all designed to span introns and were characterized previously (11). The relative concentrations of Lpin1, HPRT and TBP were determined by plotting the threshold cycle (Ct) versus the log of the serial dilution points, and Lpin1 levels determined after normalizing to endogenous controls. Relative Lpin1 mRNA levels are expressed as the ratio of Lpin1 RNA to the endogenous controls.
Owing to non-normal distributions, the Spearman correlation was calculated to assess the relationships of glucose, insulin and HOMA-IR with murine Lpin1 expression levels.
Expression data on the LPIN1 gene in human adipose tissue were obtained from a previous whole-genome expression study of fat biopsies from 19 individuals selected from families ascertained for FCHL and low HDL-C (23). TheHuman Genome U133A array chip used in the previous study included three probe sets for the LPIN1 gene (212272_at, 212274_at and 212276_at). In this study, to avoid the possible bias introduced through incomplete reverse transcription due to the relatively large size of the LPIN1 transcript (5.34 kb), we used the probe set 212276_at located in the 3' end of the gene and ignored the two more upstream located probe sets. The Spearman correlation was calculated to examine the relationship between the normalized expression levels of LPIN1 and BMI, serum TG, glucose and insulin levels.
Study subjects for LPIN1 genotype analysis
Finnish low HDL-C and FCHL families.
A total of 1109 individuals from 92 dyslipidemic families were genotyped for association analyses. The study sample consisted of 426 individuals from 39 Finnish low HDL-C families and 683 individuals from 53 Finnish FCHL families (24–27) (Table 1). Fat biopsies were obtained from 19 family members (eight males, 11 females). Clinical characteristics and the distribution of tested quantitative traits for the entire study sample, as well as for males and females taken separately, are presented in Table 1. Each study subject provided a written informed consent prior to participating in the study. All samples were collected in accordance with the Helsinki declaration, and the ethics committees of the participating centers approved the study design.
Obesity case–control study sample.
The study sample consisted of 1298 Finnish subjects of which 477 were cases (218 females/259 males) and 821 controls (367/454) (Table 4). All the cases had BMI 
30 kg/m2 and the controls 20–25 kg/m2. In detail, 279 obese Finnish cases were recruited from three obesity clinics [Helsinki University Hospital (28), Peijas Hospital and Tampere University Hospital]; 53 obese cases from the Finnish twin cohort (29) and 145 obese cases from the National FINRISK97 cohort in Finland (30). These patients had contacted the obesity clinic on their own initiative and when first contacting the clinic, the patients had a BMI 35 kg/m2. The obese cases from the Finnish twin cohort had a BMI30 kg/m2 and an obese sibling (BMI30 kg/m2) in the family. The obese cases selected from the National FINRISK97 cohort had a BMI 35 kg/m2. All lean control subjects (n=821) were recruited from the same geographical regions as the obese cases (Table 4).
Biochemical analyses
Serum TG, glucose and insulin levels were measured in the Finnish FCHL and low HDL-C families as described earlier (27,31). The homeostasis model for insulin resistance index (HOMA-IR) was calculated according to the formula: [fasting glucose (mmol/l)xfasting insulin (µU/ml)]/22.5. The post-heparin plasma activities of LPL and HL were determined as previously (32).
Marker selection and genotyping
Public, NCBI (www.ncbi.nlm.nih.gov) and commercial, Celera Genomics (www.celera.com), databases were used to find SNPs in the LPIN1 gene located on short arm of chromosome 2 (2p21). To confirm the polymorphic character of the selected SNPs, 16 Finnish control individuals were tested. If no heterozygotes were detected or when adjacent SNPs were observed in complete LD, the particular SNP was discarded from further studies. The physical order of SNP markers was determined according to the July 2003 version of UCSC database (http://genome.ucsc.edu/).
SNP genotyping was performed using the homogenous mass extension reaction on the Mass Array System (Sequenom, San Diego, CA, USA), according to the manufacturer's instructions. All the seven SNPs were in Hardy–Weinberg equilibrium in the lean control population.
Statistical analyses
Association of LPIN1 markers in families.
Pairwise LD among markers was assessed using ldmax option in the GOLD (graphical overview of LD) program (33). This program uses the expectation–maximization algorithm to estimate haplotype frequencies in a population when phase is unknown (34).
Preferential transmission of LPIN1 SNPs and their haplotypes to quantitative lipid traits in offspring was tested using the FBAT and haplotype based association test software (35) with options optimize offset (-o) to obtain information from the entire families and haplotype permutation (-p) in FCHL and low HDL-C families. Permutation was done 100 000 times to obtain empiric P-values for the haplotype association analyses. Prior to these analyses, BMI, TG, glucose and insulin levels and LPL and HL activities were log-transformed when necessary and adjusted for age and sex by multiple regression analysis, and the residuals were used in the analyses. When analyzing males and females separately, the regressions to adjust for age were conducted for each sex separately.
Association of LPIN1 SNPs in cases and controls.
To test for differences in SNP frequencies between cases and controls, a chi-squared test of association was performed for each SNP.
The quantitative trait BMI was tested for association with each of the LPIN1 SNPs using a measured genotype approach. As there were differences in BMI distributions of females and males within the obese and lean groups, the measured genotype analyses were performed separately in four groups; obese males, obese females, lean males and lean females. Because BMI did not follow a normal distribution in any of these groups, the non-parametric Kruskal–Wallis test was used toevaluate differences in BMI ranks categorized by LPIN1 SNP genotypes. BMI values were corrected for age before ranking the residuals. These analyses were carried out using the SPSS 12.0.1 software (SPSS Inc., Lead Technologies, Chicago, IL, USA).
|
ACKNOWLEDGEMENTS |
|---|
We wish to warmly thank the patients for participating in this study. Dr Pertti Mustajoki and Dr Jorma Salmi are greatly appreciated for patient collection. Minna Levander is thanked for excellent technical assistance and Kaisa Silander for Sequenom expertise. This study was supported by the Finnish Cultural Foundation, Sigrid Jusélius Foundation, Finnish Heart Foundation, The Center of Excellence in Disease Genetics of the Academy of Finland, the European Commission (BM4-CT95-0662 and QLG2-CT-2002-01254), Biomedicum Helsinki Foundation, United States Public Health Service grants HL28481 and HL-70150, American Heart Association Grant 0430180 N, UCLA Medical Scientist Training Program grant GM08042, The Helsingin Sanomat Centennial Foundation.
Conflict of Interest statement. The authors have declared that no conflict of interest exists.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
- Kopelman, P.G. (2000) Obesity as a medical problem. Nature, 404, 635–643.[Medline]
- International Agency for Research on Cancer (IARC) (2002) IARC Handbooks of Cancer Prevention. Lyon, France, Vol. 6.
- Frayn, K.N. (2002) Adipose tissue as a buffer for daily lipid flux. Diabetologia, 45, 1201–1210.[CrossRef][ISI][Medline]
- Reue, K. and Peterfy, M. (2000) Mouse models of lipodystrophy. Curr.Atheroscler. Rep., 2, 390–396.
- Robinson, W., Dinulescu, D.M. and Cone, R.D. (2000) Genetic models of obesity and energy balance in the mouse. Annu. Rev. Genet., 34, 687–745.[CrossRef][ISI][Medline]
- Nawrocki, A.R. and Scherer, P.E. (2004) The delicate balance between fat and muscle: adipokines in metabolic disease and musculoskeletal inflammation. Curr. Opin. Pharmacol., 4, 281–289.[CrossRef][ISI][Medline]
-
Unger, R.H. (2003) Minireview: weapons of lean body mass destruction: the role of ectopic lipids in the metabolic syndrome. Endocrinology, 144, 5159–5165.
[Abstract/Free Full Text] - Peterfy, M., Phan, J., Xu, P. and 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.[CrossRef][ISI][Medline]
-
Reue, K., Xu, P., Wang, X.P. and Slavin, B.G. (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.
[Abstract/Free Full Text] -
Langner, C.A., Birkenmeier, E.H., Ben-Zeev, O., Schotz, M.C., Sweet,H.O., Davisson, M.T. and Gordon, J.I. (1989) The fatty liver dystrophy (fld) mutation. A new mutant mouse with a developmental abnormality in triglyceride metabolism and associated tissue-specific defects in lipoprotein lipase and hepatic lipase activities. J. Biol. Chem., 264, 7994–8003.
[Abstract/Free Full Text] -
Phan, J., Peterfy, M. and Reue, K. (2004) Lipin expression preceding peroxisome proliferator-activated receptor-gamma is critical for adipogenesis in vivo and in vitro. J. Biol. Chem., 279, 29558–29564.
[Abstract/Free Full Text] - Phan, J., Peterfy, M. and Reue, K. (2005) Biphasic expression of lipin suggests dual roles in adipocyte development. Drug News Perspect., 18, 5–11.[CrossRef][ISI][Medline]
- Surwit, R.S., Kuhn, C.M., Cochrane, C., McCubbin, J.A. and Feinglos,M.N. (1988) Diet-induced type II diabetes in C57BL/6J mice. Diabetes, 37, 1163–1167.[Abstract]
-
Burcelin, R., Crivelli, V., Dacosta, A., Roy-Tirelli, A. and Thorens, B. (2002) Heterogeneous metabolic adaptation of C57BL/6J mice to high-fat diet. Am. J. Physiol. Endocrinol. Metab., 282, E834–E842.
[Abstract/Free Full Text] - Cao, H. and Hegele, R.A. (2002) Identification of single-nucleotide polymorphisms in the human LPIN1 gene. J. Hum. Genet., 47, 370–372.[CrossRef][ISI][Medline]
- Comuzzie, A.G., Hixson, J.E., Almasy, L., Mitchell, B.D., Mahaney, M.C., Dyer, T.D., Stern, M.P., MacCluer, J.W. and Blangero, J. (1997) A majorquantitative trait locus determining serum leptin levels and fat massis located on human chromosome 2. Nat. Genet., 15, 273–276.[CrossRef][ISI][Medline]
- Hager, J., Dina, C., Francke, S., Dubois, S., Houari, M., Vatin, V., Vaillant, E., Lorentz, N., Basdevant, A., Clement, K. et al. (1998) Agenome-wide scan for human obesity genes reveals a major susceptibility locus on chromosome 10. Nat. Genet., 20, 304–308.[CrossRef][ISI][Medline]
- Rotimi, C.N., Comuzzie, A.G., Lowe, W.L., Luke, A., Blangero, J. and Cooper, R.S. (1999) The quantitative trait locus on chromosome 2 for serum leptin levels is confirmed in African-Americans. Diabetes, 48, 643–644.[ISI][Medline]
-
Huffman, T.A., Mothe-Satney, I. and Lawrence, J.C. Jr (2002) Insulin-stimulated phosphorylation of lipin mediated by the mammalian target of rapamycin. Proc. Natl Acad. Sci. USA, 99, 1047–1052.
[Abstract/Free Full Text] - Fingar, D.C. and Blenis, J. (2004) Target of rapamycin (TOR): an integrator of nutrient and growth factor signals and coordinator of cell growth and cell cycle progression. Oncogene, 23, 3151–3171.[CrossRef][ISI][Medline]
-
Lan, H., Rabaglia, M.E., Stoehr, J.P., Nadler, S.T., Schueler, K.L., Zou,F., Yandell, B.S. and Attie, A.D. (2003) Gene expression profiles ofnondiabetic and diabetic obese mice suggest a role of hepatic lipogeniccapacity in diabetes susceptibility. Diabetes, 52, 688–700.
[Abstract/Free Full Text] -
Turner, M.J., Kleeberger, S.R. and Lightfoot, J.T. (2005) Influence of genetic background on daily running-wheel activity differs with aging. Physiol. Genomics, 22, 76–85.
[Abstract/Free Full Text] -
Naukkarinen, J., Gentile, M., Soro-Paavonen, A., Saarela, J., Koistinen, H.A., Pajukanta, P., Taskinen, M.R., Peltonen, L. (2005) USFI and dyslipidemias: converging evidence for a functional intronic variant. Hum. Mol. Genet., 14, 2595–2605.
[Abstract/Free Full Text] - Pajukanta, P., Terwilliger, J.D., Perola, M., Hiekkalinna, T., Nuotio, I., Ellonen, P., Parkkonen, M., Hartiala, J., Ylitalo, K., Pihlajamäki, J. et al. (1999) Genomewide scan for familial combined hyperlipidemia genes in Finnish families, suggesting multiple susceptibility loci influencing triglyceride, cholesterol, and apolipoprotein B levels. Am. J. Hum. Genet., 64, 1453–1463.[CrossRef][ISI][Medline]
- Lilja, H.E., Soro, A., Ylitalo, K., Nuotio, I., Viikari, J.S., Salomaa, V., Vartiainen, E., Taskinen, M.R., Peltonen, L. and Pajukanta, P. (2002) Acandidate gene study in low HDL-cholesterol families provides evidence for the involvement of the APOA2 gene and the APOA1C3A4 gene cluster. Atherosclerosis, 164, 103–111.[CrossRef][ISI][Medline]
- Soro, A., Pajukanta, P., Lilja, H.E., Ylitalo, K., Hiekkalinna, T., Perola,M., Cantor, R.M., Viikari, J.S., Taskinen, M.R. and Peltonen, L. (2002) Genome scans provide evidence for low-HDL-C loci on chromosomes 8q23, 16q24.1–24.2, and 20q13.11 in Finnish families. Am. J. Hum. Genet., 70, 1333–1340.[CrossRef][ISI][Medline]
-
Lilja, H.E., Suviolahti, E., Soro-Paavonen, A., Hiekkalinna, T., Day, A., Lange, K., Sobel, E., Taskinen, M.R., Peltonen, L., Perola, M. and Pajukanta, P. (2004) Locus for quantitative HDL-cholesterol on chromosome 10q in Finnish families with dyslipidemia. J. Lipid Res., 45, 1876–1884.
[Abstract/Free Full Text] - Oksanen, L., Mustajoki, P., Kaprio, J., Kainulainen, K., Jänne, O., Peltonen, L. and Kontula, K. (1996) Polymorphism of the beta 3-adrenergic receptor gene in morbid obesity. Int. J. Obes. Relat. Metab. Disord., 20, 1055–1061.[ISI][Medline]
-
Öhman, M., Oksanen, L., Kaprio, J., Koskenvuo, M., Mustajoki, P., Rissanen, A., Salmi, J., Kontula, K. and Peltonen, L. (2000) Genome-wide scan of obesity in Finnish sibpairs reveals linkage to chromosome Xq24. J. Clin. Endocrinol. Metab., 85, 3183–3190.
[Abstract/Free Full Text] -
Vartiainen, E., Jousilahti, P., Alfthan, G., Sundvall, J., Pietinen, P. andPuska, P. (2000) Cardiovascular risk factor changes in Finland, 1972–1997. Int. J. Epidemiol., 29, 49–56.
[Abstract/Free Full Text] -
Soro, A., Jauhiainen, M., Ehnholm, C. and Taskinen, M.R. (2003) Determinants of low HDL levels in familial combined hyperlipidemia. J. Lipid Res., 44, 1536–1544.
[Abstract/Free Full Text] - Huttunen, J.K., Ehnholm, C., Kinnunen, P.K. and Nikkilä, E.A. (1975) An immunochemical method for the selective measurement of two triglyceride lipases in human postheparin plasma. Clin. Chim. Acta, 63, 335–347.[CrossRef][Medline]
-
Abecasis, G.R. and Cookson, W.O. (2000) GOLD—graphical overview of linkage disequilibrium. Bioinformatics, 16, 182–183.
[Abstract/Free Full Text] - Excoffier, L. and Slatkin, M. (1995) Maximum-likelihood estimation of molecular haplotype frequencies in a diploid population. Mol. Biol. Evol., 12, 921–927.[Abstract]
-
Laird, N.M., Horvath, S. and Xu, X. (2000) Implementing a unified approach to family-based tests of association. Genet. Epidemiol., 19(Suppl. 1), S36–S42.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. A. Fawcett, N. Grimsey, R. J.F. Loos, E. Wheeler, A. Daly, M. Soos, R. Semple, H. Syddall, C. Cooper, S. Siniossoglou, et al. Evaluating the Role of LPIN1 Variation in Insulin Resistance, Body Weight, and Human Lipodystrophy in U.K. Populations Diabetes, September 1, 2008; 57(9): 2527 - 2533. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Zhang, L. O'Loughlin, D. N. Brindley, and K. Reue Regulation of lipin-1 gene expression by glucocorticoids during adipogenesis J. Lipid Res., July 1, 2008; 49(7): 1519 - 1528. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Donkor, L. M. Sparks, H. Xie, S. R. Smith, and K. Reue Adipose Tissue Lipin-1 Expression Is Correlated with Peroxisome Proliferator-Activated Receptor {alpha} Gene Expression and Insulin Sensitivity in Healthy Young Men J. Clin. Endocrinol. Metab., January 1, 2008; 93(1): 233 - 239. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Wiedmann, M. Fischer, M. Koehler, K. Neureuther, G. Riegger, A. Doering, H. Schunkert, C. Hengstenberg, and A. Baessler Genetic Variants Within the LPIN1 Gene, Encoding Lipin, Are Influencing Phenotypes of the Metabolic Syndrome in Humans Diabetes, January 1, 2008; 57(1): 209 - 217. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Reue and J. Donkor Genetic Factors in Type 2 Diabetes: All in the (Lipin) Family Diabetes, December 1, 2007; 56(12): 2842 - 2843. [Full Text] [PDF]
|
|
|
|
|
|
Y. S. Aulchenko, J. Pullen, W. P. Kloosterman, M. Yazdanpanah, A. Hofman, N. Vaessen, P. J.L.M. Snijders, D. Zubakov, I. Mackay, M. Olavesen, et al. LPIN2 Is Associated With Type 2 Diabetes, Glucose Metabolism, and Body Composition Diabetes, December 1, 2007; 56(12): 3020 - 3026. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Croce, J. C. Eagon, L. L. LaRiviere, K. M. Korenblat, S. Klein, and B. N. Finck Hepatic Lipin 1{beta} Expression Is Diminished in Insulin-Resistant Obese Subjects and Is Reactivated by Marked Weight Loss Diabetes, September 1, 2007; 56(9): 2395 - 2399. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Yatsu, N. Mizuki, N. Hirawa, A. Oka, N. Itoh, T. Yamane, M. Ogawa, T. Shiwa, Y. Tabara, S. Ohno, et al. High-Resolution Mapping for Essential Hypertension Using Microsatellite Markers Hypertension, March 1, 2007; 49(3): 446 - 452. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Donkor, M. Sariahmetoglu, J. Dewald, D. N. Brindley, and K. Reue Three Mammalian Lipins Act as Phosphatidate Phosphatases with Distinct Tissue Expression Patterns J. Biol. Chem., February 9, 2007; 282(6): 3450 - 3457. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. van Harmelen, M. Ryden, E. Sjolin, and J. Hoffstedt A role of lipin in human obesity and insulin resistance: relation to adipocyte glucose transport and GLUT4 expression J. Lipid Res., January 1, 2007; 48(1): 201 - 206. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Xu, W.N. P. Lee, J. Phan, M. F. Saad, K. Reue, and I. J. Kurland Lipin Deficiency Impairs Diurnal Metabolic Fuel Switching Diabetes, December 1, 2006; 55(12): 3429 - 3438. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J van Erk, W. A. Blom, B. van Ommen, and H. F. Hendriks High-protein and high-carbohydrate breakfasts differentially change the transcriptome of human blood cells. Am. J. Clinical Nutrition, November 1, 2006; 84(5): 1233 - 1241. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Yao-Borengasser, N. Rasouli, V. Varma, L. M. Miles, B. Phanavanh, T. N. Starks, J. Phan, H. J. Spencer III, R. E. McGehee Jr., K. Reue, et al. Lipin Expression Is Attenuated in Adipose Tissue of Insulin-Resistant Human Subjects and Increases With Peroxisome Proliferator-Activated Receptor {gamma} Activation. Diabetes, October 1, 2006; 55(10): 2811 - 2818. [Abstract] [Full Text] [PDF]</ |