Identification of ALOX5 as a gene regulating adiposity and pancreatic function
- First Online:
We previously used an integrative genetics approach to demonstrate that 5-lipoxygenase (5-LO) deficiency in mice (Alox5−/−) protects against atherosclerosis despite increasing lipid levels and fat mass. In the present study, we sought to further examine the role of 5-LO in adiposity and pancreatic function.
Alox5−/− and wild-type (WT) mice were characterised with respect to adiposity and glucose/insulin metabolism using in vivo and in vitro approaches. The role of ALOX5 in pancreatic function in human islets was assessed through short interfering RNA (siRNA) knockdown experiments.
Beginning at 12 weeks of age, Alox5−/− mice had significantly increased fat mass, plasma leptin levels and fasting glucose levels, but lower fasting insulin levels (p < 0.05). Although Alox5−/− mice did not exhibit insulin resistance, they had impaired insulin secretion in response to a bolus glucose injection. Histological analyses revealed that Alox5−/− mice had increased islet area, beta cell nuclear size, and numbers of beta cells/mm2 islet (p < 0.05), indicative of both hyperplasia and hypertrophy. Basal and stimulated insulin secretion in isolated Alox5−/− islets were significantly lower than in WT islets (p < 0.05) and accompanied by a three- to fivefold decrease in the expression of the genes encoding insulin and pancreatic duodenal homeobox 1 (Pdx1). Direct perturbation of ALOX5 in isolated human islets with siRNA decreased insulin and PDX1 gene expression by 50% and insulin secretion by threefold (p < 0.05).
These results provide strong evidence for pleiotropic metabolic effects of 5-LO on adiposity and pancreatic function and may have important implications for therapeutic strategies targeting this pathway for the treatment of cardiovascular disease.
KeywordsBeta cell function Insulin secretion Leukotrienes 5-Lipoxygenase Obesity Inflammation Type 2 diabetes
intraperitoneal glucose tolerance test
intraperitoneal insulin tolerance test
Krebs Ringer bicarbonate
nuclear magnetic resonance
pancreatic duodenal homeobox 1
short interfering RNA
The 5-lipoxygenase (5-LO) pathway, which generates proinflammatory leukotrienes (LTs) from arachidonic acid, has recently attracted a great deal of attention for its potential role in traits related to cardiovascular disease (CVD). This stems from a series of recently conducted studies, which have collectively provided strong evidence for the proatherogenic role of LTs [1, 2]. Importantly, multiple lines of evidence support this concept, including biochemical, genetic and pharmacological studies in both mice and humans . For example, we demonstrated that deficiency of the gene encoding 5-lipoxygenase (Alox5) in mice protects against the development of aortic lesions [3, 4], and that promoter variants of the human ALOX5 gene are associated with increased carotid intima–media thickness . Other groups have also reported findings that support the involvement of the 5-LO/LT pathway in CVD traits [6, 7, 8, 9, 10, 11, 12, 13].
More recently, we used an integrative genomics approach with inbred mouse strains, gene-targeted mice and microarray profiling to demonstrate that 5-LO also contributes to metabolic phenotypes, such as body fat, bone density and lipid levels . These observations raise important questions regarding the potential pleiotropic effects of the 5-LO pathway on components of the metabolic syndrome. To further explore these inter-relationships, we characterised Alox5−/− mice for traits related to obesity and type 2 diabetes.
Alox5−/− mice were purchased from Jackson Laboratories (Bar Harbor, ME, USA) and backcrossed onto a C57BL/6J (B6) background for more than ten generations. Wild-type (WT) B6 mice were either bred in house or purchased from Jackson Laboratories. Since no differences were observed between WT mice from either source, the data from both types of control mice were combined. All animals were housed four per cage at 25°C on a 10 h dark/14 h light cycle and maintained on a chow diet (No. 5015; Purina, Richmond, IN, USA). All procedures were in accordance with the current National Research Council Guide for the care and use of laboratory animals, and were approved by the UCLA Animal Research Committee.
Mice were bled retro-orbitally under isoflurane anaesthesia or through the tail vein while restrained and conscious. Glucose levels were determined using commercially available kits from Sigma (St Louis, MO, USA) or the OneTouch Ultra Blood Glucose Monitoring System (LifeScan, Milpitas, CA, USA). We previously established the accuracy of glucose values using this glucometer compared with the Sigma kit (M. Mehrabian, unpublished data). Insulin and leptin levels were measured using commercial ELISA kits from Crystal Chemical (Chicago, IL, USA) and R&D Systems (Minneapolis, MN, USA), respectively. All measurements were performed in duplicate or triplicate according to the manufacturer’s instructions.
Whole body fat, fluids and lean tissue mass of isoflurane-anaesthetised mice were determined using a Bruker Optics Minispec nuclear magnetic resonance (NMR) analyser (The Woodlands, TX, USA) according to the manufacturer’s recommendations. The mass of individual fat depots (retroperitoneal, epididymal, subcutaneous, and omental) was determined by dissecting and weighing each fat pad separately after the mice were killed.
Intraperitoneal glucose and insulin tolerance tests
Intraperitoneal insulin tolerance tests (IPITTs) were performed by administering an intraperitoneal injection (0.75 U/kg body weight) of recombinant human insulin (Novolin; Novo Nordisk, Bagsværd, Denmark) to conscious mice fasted for 5 h. Plasma glucose levels were measured in blood samples obtained through the tail vein using a glucometer, as described above. For intraperitoneal glucose tolerance tests (IPGTTs), mice were fasted for 12 h overnight and injected with a bolus (1 mg/g body weight) of glucose (10% [wt/vol.] in sterile H2O) into the peritoneal cavity. Blood samples were obtained from conscious mice through the tail vein or from isoflurane-anaesthetised mice through the retro-orbital plexus at 0, 15, 30, 60, 90 and 120 min post-injection. Plasma glucose and insulin levels were measured as described above.
After the animals were killed, the pancreas was dissected, weighed, fixed in formalin and embedded in paraffin. Tissue sections were deparaffinised, rehydrated and incubated with guinea pig anti-insulin antibody (Dako, Glostrup, Denmark) followed by detection with a fluorescein-conjugated rabbit anti-guinea pig antibody (Jackson ImmunoResearch Laboratories, West Grove, PA, USA). The samples were subsequently labelled for glucagon with rabbit anti-glucagon or with a mixture of rabbit anti-somatostatin and anti-pancreatic polypeptide antibodies (Dako), followed by detection with donkey anti-rabbit indocarbocyanine (Cy3)-conjugated antibody (Jackson ImmunoResearch Laboratories). Beta cell mass was analysed using Openlab software (Improvision, Waltham, MA, USA). The relative area of beta cells (green fluorescence) was determined by calculation of the cross-sectional beta cell area divided by the cross-sectional area of total tissue. The beta cell mass per pancreas was estimated as the product of the relative cross-sectional area of beta cells per total tissue and the weight of the pancreas. Beta cell number per mm2 of islet, nuclear size, nuclear distance and islet size were quantified using an inverted system microscope (IX70; Olympus America, Melville, NY, USA) and Image-Pro Plus software (Media Cybernetics, Silver Springs, MD, USA). The number of beta cells per islet was calculated by multiplying islet area in mm2 by the number of beta cells per mm2. Ten representative 20 µm sections from each pancreas (spanning the width of the pancreas) were used in these analyses, and the results are from ten male mice of each genotype, from three litters, at 12 weeks of age.
Islet isolation and culture
Human islets were isolated from the pancreas of five healthy organ donors at the University of Illinois at Chicago, as described previously  All donors gave informed consent, and institutional review board approval was obtained from the participating institutions. Mouse islets were isolated by bile duct perfusion and collagenase digestion, as described elsewhere . The islets were cultured on matrix-coated plates derived from bovine corneal endothelial cells (Novamed, Jerusalem, Israel), allowing the cells to attach to the dishes and spread . Human islets were cultured in CMRL 1066 medium containing 5.5 mmol/l glucose, and mouse islets in RPMI 1640 medium containing 11.1 mmol/l glucose, both supplemented with 100 U/ml penicillin, 100 µg/ml streptomycin, and 10% FCS (Invitrogen, Carlsbad, CA, USA).
Insulin release and content measurements
Chronic insulin release over 48 h was evaluated in the culture medium collected before the termination of the experiment. For acute insulin release in response to glucose, mouse islets or transfected human islets were washed and pre-incubated for 30 min in Krebs Ringer bicarbonate buffer (KRB) containing 2.8 mmol/l glucose and 0.5% BSA. KRB was then replaced with KRB containing 2.8 mmol/l glucose for 60 min (basal), followed by an additional 60 min in KRB with 16.7 mmol/l glucose (stimulated). Thereafter, islets were incubated in 1 ml 0.18 mol/l HCl in 70% ethanol and left overnight at 4°C. Insulin was determined using a mouse insulin ELISA kit (ALPCO Diagnostics, Salem, NH, USA). The stimulatory index was calculated by dividing stimulated insulin secretion levels by basal insulin secretion. Beta cell area of the stimulated islets was calculated under the microscope using Image-Pro Plus software. All experiments were performed in three independent rounds using islets isolated from three litters of 12 week old male Alox5−/− and WT mice (n = 10 for each strain), respectively, with three dishes for each genotype group and 20 islets per dish.
RNA interference experiments
Short interfering RNAs (siRNAs) designed for the specific knockdown of ALOX5 expression or as a non-specific control were synthesised by Ambion (Austin, TX, USA). Isolated human islets were plated on extracellular matrix-coated dishes and cultured in medium containing 5.5 mmol/l glucose for 24 h. The islets were then transfected with 100 nmol/l siRNA using 2.5 µl/ml Lipofectamine 2000 in OptiMEM media (Invitrogen) according to the manufacturer’s protocols. The transfection medium was replaced after 12 h with antibiotic-free culture media, and 96 h later the islets were harvested for analysis. All experiments were performed in triplicate with three dishes for each treatment group, with 20 islets per dish from five donors in five independent rounds.
RNA extraction and real-time quantification
Total RNA was extracted from adipose tissue or pancreatic islets using RNeasy Mini kits (Qiagen, Valencia, CA, USA). RNA was isolated from 50 mouse islets for each genotype group, which were obtained from ten 12 week old mice, whereas 100 islets from each treatment group were used to extract RNA for the human islet experiments. The cDNA was prepared from 1 µg of total RNA using reverse transcription kits (Life Technologies, Gaithersburg, MD, USA; Applied Biosystems [ABI], Foster City, CA, USA; or Invitrogen) according to the manufacturers’ protocols. Real-time expression assays were performed either on the Light Cycler quantitative PCR system (Roche, Basel, Switzerland) or an ABI 7900HT instrument. Real-time reactions were carried out in triplicate with either SYBR Green assays, as described previously [18, 19], or pre-developed assays from ABI . For the SYBR Green assays, the RNA levels in each sample were calculated using standard curves generated from serial tenfold dilutions of pooled RNA from WT mice. For the pre-developed assays, RNA levels in each sample were calculated using pooled WT mice as the reference sample. Each sample was normalised to endogenous controls, such as GAPDH/Gapdh, ACTB, or TUBA1B/Tuba1b, and the replicates were averaged to determine differences between Alox5−/− and WT mice or between human islets treated with control or ALOX5-specific siRNAs. A list of the primer sequences used for the SYBR Green assays is provided in Electronic supplementary material Table 1.
Differences in measured variables between Alox5−/− and WT mice or the siRNA treatments in human islets were determined by Student’s t test using Statview, version 5.0 (SAS Institute, Cary, NC, USA). Values are expressed as mean ± SEM, and differences were considered statistically significant at p < 0.05.
Increased adiposity in Alox5−/− mice
Measures of body weight and adiposity in Alox5−/− and WT mice
Percent body fat
Percent body fat
19.9 ± 0.3
2.5 ± 0.3
12.4 ± 1.2
26.0 ± 2.3*
2.8 ± 0.2*
10.8 ± 0.5*
18.6 ± 0.1
2.0 ± 0.3
12.2 ± 1.3
23.8 ± 1.7
1.9 ± 0.4
7.9 ± 1.0
24.1 ± 0.4*
2.8 ± 0.8*
11.7 ± 1.4*
29.0 ± 1.6*
3.5 ± 0.2*
12.2 ± 0.2*
21.6 ± 0.3
2.0 ± 0.1
9.5 ± 0.5
26.4 ± 1.7
2.0 ± 0.4
7.5 ± 0.9
27.5 ± 0.2*
2.9 ± 0.6*
10.6 ± 0.24*
30.0 ± 1.7*
3.9 ± 0.4*
12.9 ± 0.5*
22.8 ± 0.1
2.0 ± 0.01
9.3 ± 0.2
26.4 ± 1.5
2.0 ± 0.1
7.6 ± 0.2
28.7 ± 0.1*
3.4 ± 0.4*
11.9 ± 1.3*
31.2 ± 1.9*
3.9 ± 0.5*
12.4 ± 1.2*
23.4 ± 0.2
2.0 ± 0.2
8.8 ± 0.9
27.5 ± 1.5
2.0 ± 0.2
7.3 ± 0.7
Impaired glucose and insulin metabolism in Alox5−/−mice
To explore these differences further, we first assessed insulin sensitivity by IPGTT. As shown in Fig. 2d, 12-week-old male and female Alox5−/− mice exhibited significantly higher plasma glucose levels than WT mice at all time points following a bolus glucose injection. This was accompanied by significantly lower insulin levels at the 0, 15, and 30 min time points (Fig. 2e). These experiments were repeated at 18 weeks of age, where again we observed significantly higher fasting and fed plasma glucose (Fig. 2f) and lower fasting insulin levels (Fig. 2g) in Alox5−/− mice. An IPGTT at 18 weeks of age also demonstrated significantly elevated plasma glucose levels in Alox5−/− mice at all time points post-injection (Fig. 2h). Based on these observations, we next carried out IPITTs to assess insulin sensitivity. Interestingly, although Alox5−/− mice had elevated plasma glucose levels at all time points following insulin injection, including at baseline (Fig. 2i), the rate of glucose disposal was the same as that for WT mice, since the difference between plasma glucose levels at 60 min and those at 0 min were not significantly different between the two strains (Fig. 2j). These results suggest that 5-LO deficiency does not alter insulin sensitivity noticeably, despite an increase in adiposity.
Another series of IPGTTs was carried out on mice at 23 weeks of age, to gain a better understanding of glucose and insulin dynamics. Similar to the results in younger mice, 23-week-old Alox5−/− mice exhibited higher plasma glucose levels than the WT mice at baseline and throughout the duration of the experiment (Fig. 2k). Plasma insulin levels were significantly lower in the Alox5–/– mice at baseline and at the 10, 15 and 30 min time points (Fig. 2l), thus suggesting an impairment of insulin secretion. We also used the IPGTT data to calculate the mean AUC, which is increased for glucose and decreased for insulin levels in Alox5−/− mice compared with WT mice (Fig. 2k,l).
Impaired beta cell function in Alox5−/− mice
Direct effect of 5-LO on INS expression and secretion in human islets
With the exception of the genes known to cause maturity-onset diabetes of the young, the genetic factors controlling insulin secretion are not well known. Notably, recent genetic studies in humans have provided evidence that genes controlling beta cell function also play an important role in adult-onset type 2 diabetes, suggesting that insulin secretion-related mechanisms may make more significant genetic contributions to common forms of type 2 diabetes than previously recognised [22, 23, 24, 25, 26]. Thus, identification of novel pathways regulating beta cell function and/or insulin secretion may lead to a better understanding of type 2 diabetes and its underlying aetiology. We now provide evidence that the 5-LO/LT pathway is another mechanism by which such processes are regulated. Importantly, our observations are in agreement with a series of earlier studies carried out by Metz et al. demonstrating that lipoxygenase inhibitors decrease glucose-stimulated insulin secretion in a dose-dependent manner [27, 28] and that arachidonic acid metabolites play a role as ‘third messengers’ for hormone release [29, 30]. More recently, Prasad and colleagues have shown that 12/15-LO, a related enzyme that also metabolises arachidonic acid, influences beta cell function and insulin secretion as well .
In this study, we demonstrate that 5-LO deficiency in mice impairs glucose/insulin homeostasis from an early age. For example, in vitro and in vivo experiments showed that the insulin response is decreased in isolated islets from 12-week-old Alox5−/− mice. These observations suggest that 5-LO deficiency alters beta cell function relatively early in life and prior to the onset of more pronounced obesity. Such abnormalities continue as the mice age, since decreased fasting and glucose-stimulated plasma insulin levels are also observed at 18 and 23 weeks of age. It is possible that overall pancreatic function is also compromised given the significantly reduced weight of the pancreas and insulin content of Alox5−/− islets. Interestingly, our histological analyses suggest that Alox5−/− islets are larger and contain beta cells that have undergone both hyperplasia and hypertrophy. These morphological changes may indicate compensatory mechanisms for the reduced insulin secretory capacity of Alox5−/− islets and/or that LTs are required for normal pancreatic development. Furthermore, siRNA knockdown experiments in human islets decreased gene expression and insulin secretion, consistent with the observations in isolated Alox5−/− islets. Taken together, our results indicate a direct role for 5-LO in the regulation of insulin production and secretion, both in an acute setting and over the long term.
Obesity is now widely recognised to be associated with a heightened inflammatory state involving elevations in the circulating levels and production of inflammatory cytokines such as C-reactive protein, IL-6 and TNF-α . Moreover, two separate reports have shown that adipose tissue is characterised by macrophage infiltration, which could affect its normal function and inflammatory properties [33, 34]. In this regard, we show that although adiposity is increased in Alox5−/− mice from as early as 12 weeks of age, the expression of inflammatory genes in adipose tissue is lower. Thus, the mechanism by which 5-LO and LT deficiency leads to abnormal adipose biology and increased fat mass is not entirely clear. It should be noted, however, that 5-LO could presumably play a direct mechanistic role given that it is produced in both macrophages and adipose tissue.
Leptin is secreted by adipose tissue to regulate energy expenditure and food consumption. Interestingly, Alox5−/− mice exhibit hyperleptinaemia but do not have increased food intake , suggesting that leptin signalling in the hypothalamus is either not affected by 5-LO deficiency or that there are other compensatory mechanisms. In addition, elevated plasma leptin levels in Alox5−/− mice appear to result from transcriptional upregulation of the gene, although it is possible that other mechanisms such as decreased clearance and/or leptin resistance may be involved as well. The effect of elevated leptin levels on the so-called ‘adipo-insular axis’ may also provide another potential mechanism for the impaired insulin secretion in Alox5−/− mice. For example, since leptin has been shown to inhibit insulin biosynthesis and secretion [35, 36], the increased plasma leptin in Alox5−/− mice could also decrease insulin production and secretion from beta cells, in addition to what is directly affected by 5-LO itself. This could be particularly relevant in older mice in which adiposity has substantially increased.
In previous studies, we reported that 5-LO deficiency dramatically decreases aortic lesion development in two mouse models [3, 4] and that humans carrying certain ALOX5 gene variants have increased carotid intima–media thickness . Moreover, recent studies in humans and mice have demonstrated that various pharmacological modifiers of the 5-LO pathway decrease inflammatory biomarkers and aortic lesion formation, respectively [9, 13, 37], raising the possibility of developing 5-LO pathway inhibitors for the treatment of CVD. Our results suggest that it would be important for such therapeutic strategies to carefully consider potentially undesirable metabolic effects. While complete inactivation of 5-LO may result in developmental changes altering adipose tissue metabolism and/or pancreatic function that would otherwise not occur if the 5-LO pathway is inhibited in adulthood, our siRNA experiments, designed to directly perturb ALOX5 in human islets, suggest that decreased 5-LO levels and/or activity can have an acute effect on islet function and insulin secretion. In addition, the LT pathway bifurcates to the LTB4 and cysteinyl LT branches after the generation of LTA4 by 5-LO, raising the possibility that a deficiency of both types of LTs is required for the development of metabolic abnormalities. For example, montelukast, which only blocks cysteinyl LT signalling, is widely used for the treatment of asthma but has not been reported to be associated with metabolic disturbances. Finally, since arachidonic acid can also be converted into other fatty acid derivatives by the cyclooxygenase and 12/15-LO enzymes, it is possible that the metabolic phenotypes observed for the Alox5−/− mice result from the increased production of other eicosanoids owing to shunting of arachidonic acid to these alternative pathways. For example, Goulet et al. reported that production of prostaglandins and thromboxanes is increased in macrophages from Alox5−/− mice , but it is not known whether the metabolic consequences we observe with 5-LO deficiency are due directly to the lack of LTs and/or indirectly to the increased production of other eicosanoids. Thus, additional studies will be required to determine whether pharmacological antagonism of the 5-LO pathway will provide beneficial CVD effects in humans without compromising adiposity and pancreatic function.
In conclusion, we provide strong evidence that the inflammatory 5-LO/LT pathway has pleiotropic effects on adiposity and type 2 diabetes-related traits. Specifically, decreased 5-LO production leads to generalised obesity and pancreatic dysfunction in mice, and impaired insulin secretion in human islets. These results provide a new mechanism by which metabolic abnormalities associated with obesity and type 2 diabetes can arise.
This work was supported by National Institutes of Health (NIH) grants HL079353 (H. Allayee and M. Mehrabian) and HL30568 (M. Mehrabian and A. J. Lusis), American Heart Association grant 0355031Y (M. Mehrabian), American Diabetes Association Junior Faculty Grant 706JF41 (K. Maedler), the NIH/Older Americans Independence Center at UCLA (K. Maedler), and the Larry L. Hillblom Foundation (LLHF grant no. 2005 1C to K. Maedler). F. T. Schulthess is the recipient of the Swiss National Foundation Fellowship Award. Human islets were provided through the Islet Cell Resource Consortium, administered by the Administrative and Bioinformatics Coordinating Center of the City of Hope National Medical Center and supported by the National Center for Research Resources (NCRR), the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) and the Juvenile Diabetes Research Foundation. A portion of this work was conducted in a facility constructed with support from Research Facilities Improvement Program grant no. C06 (RR10600-01, CA62528-01, RR14514-01) from the NCRR.
Duality of interest
The authors declare that there is no duality of interest associated with this manuscript.