| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
,,¶,,,1
* Center for Genomics and Bioinformatics, Karolinska Institutet, Stockholm, Sweden;
Centre of Inflammation and Metabolism, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark;
The Wenner-Gren Institute, The Arrhenius Laboratories, Stockholm University, Stockholm, Sweden;
School of Life Sciences, Heriot-Watt University, Edinburgh, Scotland, UK;
|| Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm, Sweden;
¶ Department of Pharmacology, Monash University, Clayton, Victoria, Australia; and
# Department of Biochemistry, The Scripps Research Institute, Jupiter, Florida, USA
1Correspondence: C.S., Center for Genomics and Bioinformatics, Karolinska Institutet, 171 77 Stockholm, Sweden. E-mail: camilla.scheele{at}gmail.com; J.A.T., School of Life Sciences, Heriot-Watt University, Edinburgh EH14 4AS, Scotland, U.K.. E-mail: j.timmons{at}hw.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: metabolism • inactivity • mitochondria • altered gene regulation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and great importance has been attached to recent observations that PGC-1
, a transcriptional cofactor and a master regulator of mitochondrial biogenesis, is down-regulated in the skeletal muscle of type 2 diabetes patients (2)
. PGC-1 is involved in several metabolic pathways (3
, 4)
making it a plausible "disease gene". However, skeletal muscle gene expression changes in complex diseases, such as type 2 diabetes, are difficult to interpret. For example, we have demonstrated that short term physical inactivity alone can induce similar changes in human skeletal muscle PGC-1 expression to those found in type 2 diabetes (5)
, such that the interaction between diabetes per se and inactivity merits further investigation. It is also plausible that other mitochondrially associated genes participate in providing a link between mitochondrial function, physical inactivity, obesity, and type 2 diabetes.
To identify novel candidate genes associated with one physical inactivity model (5)
we utilized a small custom cDNA array platform that was enriched with druggable (6)
protein targets. We identified a modest number of candidate genes, partly reflecting the limited performance of this particular custom array. One of the genes we identified was PTEN induced putative kinase 1 (PINK1). PINK1 is a putative serine-threonine kinase that has been linked to a recessive form of familial parkinsonism (7
8
9
10)
. Although little is known about the etiology of diabetes in Parkinson’s disease, up to 80% of the patients are claimed to have impaired glucose tolerance (11)
, while several pieces of evidence point to a link between Parkinson’s disease and mitochondrial dysfunction (12)
, an association common with diabetes. Over-expressed and endogenous PINK1 localizes to the mitochondria (8
, 10
, 13)
and very recently it has been demonstrated that PINK1 plays a critical role in Drosophila oxidative flight muscle (14
15
16)
where mutated PINK1 resulted in loss of mitochondrial architecture and muscle function and was associated with increased ROS production. Indeed, antioxidant therapy could rescue neuronal cell death induced by knockdown of PINK1 in Drosophila (17)
, thus providing a link between the function of PINK1 and a postulated role of ROS in the etiology of insulin resistance (18)
.
In addition to PINK1, there are a number of transcripts produced from the human PINK1 locus (Fig. 3)
, including Dolichyl-diphosphooligosaccharide-protein glycosyltransferase (DDOST) and a natural antisense PINK1 molecule (19)
. Interestingly, DDOST translates into the AGE-R1 scavenger receptor, a counter regulatory mechanism for advanced glycation end product (AGE)-induced oxidative stress (20
21
22)
. AGE formation is enhanced by hyperglycemia (23)
, has been linked to diabetic complications (24
25
26)
and shown to accumulate in neurodegenerative disease (27)
. Partly overlapping with the 3'-end of DDOST and partly complementary to PINK1 is naPINK1, a cis-encoded natural antisense to PINK1 that we have recently functionally verified (19)
. Given the number of potential interactions between these genes and impaired metabolism, we selected the PINK1 locus for investigation in subjects with type 2 diabetes (with and without obesity). We did this simultaneously in two insulin sensitive organs (muscle and adipose tissue) and related gene expression to measures of metabolic and physical fitness. In addition, we profiled PGC-1 and two mitochondrial genes (mtND4 and Citrate synthase) to allow for direct comparison with previously published data (2
, 28
, 29)
, and evaluated the function of PINK1 using RNA interference (RNAi) in cell based models. Gene expression changes are presented from human models of physical activity (30)
and inactivity (5
, 19)
to provide additional context to the cause-effect interpretation of the clinical data.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
32
33)
, to impose five weeks of muscle inactivity in one leg of healthy volunteers. Six healthy adult subjects, four men and two women, volunteered for this study. At the time of being recruited, the subjects did not participate in any regular training programs. A written consent was given after the procedures and risks associated with participation in the study had been explained. The experimental protocol was approved by the Institutional Review Board at the University and the study was conducted in accordance with the Declaration of Helsinki as revised in 2000. The subjects were 30–56 y, with a body fat range of 9–25% (calculated from biceps, triceps, suprailiac, and front thigh skinfolds, using a Harpenden caliper). The subjects had a body mass range of 57–91 kg and a height range of 165–185 cm. The physiological results and global changes in muscle volume from this study have been reported elsewhere (5
, 34)
. Muscle biopsies were obtained from the m Soleus before and after this intervention period, and RNA isolation, real-time quantitative PCR and calculations were carried out as described previously (5)
.
For the custom microarray analysis, total RNA was prepared as described below. Data appear in supplementary File 1. All samples were labeled and hybridized as extensively described previously (35)
. The software used for the image analysis of the arrays was SpotReader (http://www.nilesscientific.com). Data analysis was performed using the R environment and Bioconductor (36)
unless otherwise stated. The images were digitalized using Quantarray and normalized using nonprint-tip loess normalization (37)
without background subtraction as this gives higher accuracy (38)
. Significantly differentially expressed genes were identified using the Significance Analysis of Microarrays (SAM) algorithm using the paired test setting as previously discussed (39)
. The significant gene list is supplied as an Excel spreadsheet. However, due to the limitation of the arrays, this data set should only be considered pilot data.
Obesity and type 2 diabetes cohort
Using a cross-sectional case-control design, participants for this study were divided into 4 groups according to the presence or absence of a diagnosis for type 2 diabetes and according to body mass index (BMI) < 27 or >35. All subjects (n=53) were recruited by advertising in a local newspaper and information of diagnosis of type 2 diabetes was based on information from each subject respectively. To verify correct diagnosis, the World Health Organization (WHO) diagnostic criteria for diabetes were employed. Exclusion criteria were treatment with insulin, recent or ongoing infection, history of malignant disease or treatment with antiinflammatory drugs. Characteristics are given in Table 1
for the four main groups: 1) normal glucose tolerant, nonobese (n=12); 2) normal glucose tolerant, obese (n=14); 3) type 2 diabetes, nonobese, (n=13); 4) type 2 diabetes, obese (n=14). Participants were given both oral and written information about the experimental procedures before giving their written informed consent. The study was approved by the Ethical Committee of Copenhagen and Frederiksberg Communities, Denmark (j.nr (KF) 01–141/04), and performed according to the Declaration of Helsinki. The Universities of Copenhagen, Karolinska Institutet and Heriot-Watt provided additional ethics approval for all molecular analysis.
|
Clinical evaluation protocol
Participants reported between 8:00 and 10:00 a.m. into the laboratory after an overnight fast. Subjects did not take their usual medication for 24 h preceding the examination, and subjects with type 2 diabetes did not take hypoglycemic medicine for one week preceding the examination day. The abdominal and limb circumferences were measured, as were body mass and height for BMI calculations. Sphygomanometric measurement of the brachial arterial blood pressure was performed on the participants resting in supine position. Blood samples were drawn from an antecubital vein. On the same day, biopsies were obtained from muscle and adipose tissue; the subjects performed an oral glucose tolerance test (OGTT), an aerobic capacity (fitness) test and were scanned using a dual-energy X-ray absorptiometry (DEXA) whole body scanner for determining body composition.
Muscle and adipose tissue biopsies
Muscle biopsies from the patient cohort were obtained from the m vastus lateralis of m quadriceps femoris using the percutaneous needle method with suction (40)
. Prior to each biopsy, local anesthesia (lidocaine, 20 mg ml–1; SAD, Denmark) was applied to the skin and superficial fascia of the biopsy site. Adipose tissue biopsies were sampled from the abdominal subcutaneous adipose tissue by the percutaneous needle biopsy technique with suction, preceded by a subcutaneous injection of lidocaine. Visible blood contamination was carefully removed and all biopsies were frozen in liquid nitrogen and subsequently stored at –80°C until further analysis.
Aerobic capacity (fitness) test and body composition
Peak aerobic capacity was determined by the Åstrand-Ryhming indirect test of maximal oxygen uptake (VO2max) (41)
. Total body and regional fat (kg) were assessed by dual-energy X-ray absorptiometry (DEXA) GE Medical Systems lunar, Prodigy Advance). The scanner was calibrated daily and a spine phantom was scanned weekly. Software version 8.8 was used to estimate regional and total fat and lean soft tissue. No sandbags or pillows were used. Landmarks were set to separate appendages (upper and lower limbs) from truncal regions: (1)
the upper limb-region was separated from trunk by a line extended from the head of humerus and the glenoid fossa of the scapula; (2)
The legs consisted of the tissue extended from the inferior border of the ischial tuborosity to the distal tip of the toes. Lean soft tissue is directly correlated to skeletal muscle mass and appendicular skeletal muscle mass (ASM) is calculated as total muscle mass in upper- and lower-limbs, which is directly correlated to total body skeletal muscle mass (42)
. Relative truncal fat is calculated as truncal fat divided by total body fat (43
, 44)
and used as an estimate of fat distribution, assumed to correlate to the amount of visceral adipose tissue (45)
. Self reported physical activity was determined as previously discussed (46)
.
Blood analyses
Plasma was obtained by drawing blood samples into glass tubes containing EDTA, and serum was obtained by drawing blood into glass tubes containing a clot-inducing plug. The tubes were immediately spun at 3500 g for 15 min at 4°C and the supernatant was isolated and stored at –20°C until analyses were performed. Plasma insulin were analyzed by radioimmunoassay (Insulin RIA 100, Amersham Pharmacia Biotech, Uppsala, Sweden) and plasma glucose was determined using an automatic analyzer (Cobas Fara, Roche, France) both as described previously (47)
. All samples and standards were run as duplicates, and the mean of duplicates was used in the statistical analyses.
Oral glucose tolerance test (OGTT)
Blood samples were drawn before, and 1 and 2 h after, drinking 500 ml of water containing 75 g of dissolved glucose. The WHO diagnostic criteria were applied (48)
. Normal glucose tolerance (NGT) was defined as fasting venous plasma glucose 
6 mM and <7.8 mM 2 h after the oral glucose load, and type 2 diabetes as fasting venous plasma glucose 
7.0 mM or venous plasma glucose 11.1 mM 2 h after the oral glucose load. HOMA2-IR was determined according to current guidelines (49
, 50)
.
Cell culture and transfection
All cell culture reagents were purchased from GIBCO (Invitrogen, Carlsbad, CA, USA) unless otherwise stated. The neuroblastoma cell line SK-N-MC was subcultured in minimal essential media (MEM) + Earle’s supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 1 mM sodium pyruvate, nonessential amino acids, 100 U/ml penicillin, and 100 µg/ml streptomycin. The neuroblastoma cell line SH-SY5Y was subcultured in DMEM/F12 supplemented with 10% FBS, nonessential amino acids, 100 U/ml penicillin and 100 µg/ml streptomycin. Cell cultures were maintained at 37°C in 5% CO2. Transfection was performed using 0.25% Lipofectamine 2000 (Invitrogen) and 20 nM siRNA, according to the manufacturer’s protocol. Following 48 h, cells were harvested for RNA isolation or, following 72 h, used for glucose uptake. PINK1 was knocked down with either PINK1-siRNA-1 (n=9) or PINK1-siRNA-2 or-3 (n=9). The siRNAs targeting PINK1 (Table S1) were prevalidated or predesigned siRNAs from Ambion (Austin, TX, USA; PINK1-siRNA-1: ID #1199; PINK1-siRNA-2: ID #1294 and PINK1-siRNA-3: ID#103456) and thus designed and chemically synthesized by Ambion. The control siRNA was predesigned by Ambion to not target any gene in the genome (Silencer®Negative Control#1, Ambion).
Primary brown preadipocyte cell culture and differentiation: Male NMRI mice (age 3–4 wk; B&K, Stockholm, Sweden) were sacrificed by CO2, and brown adipose tissue (BAT) (from the interscapular, cervical and axillary depots) was isolated as described previously (51)
. Briefly, minced tissues were digested in a collagenase (type II, Sigma Aldrich, Stockholm, Sweden) containing buffer, for 30 min at 37°C. The cell suspension was filtered and kept on ice for 20 min. After discarding the top layer (mature adipocytes) the suspension was filtered and washed in Dulbecco’s modified Eagle’s media (DMEM) and resuspended in 0.5 ml culture media per animal. Cells were cultured in 6-well plates (Falcon, BD Biosciences, Erembodegem, Belgium) in DMEM supplemented with 10% (v/v) newborn calf serum (HyClone, Erembodegem, Belgium), 2.4 nM insulin, 25 µg/ml sodium ascorbate, 10 mM HEPES, 4 mM glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin. Culture media (1.8 ml) was added to each well before 0.2 ml of cell suspension was added.
Following 24 h in culture, cells were transfected using 0.1% Lipofectamine 2000 (Invitrogen) and 20 nM short interfering RNA (siRNA) in 2.5 ml newborn calf serum-containing cell culture media (without penicillin/streptomycin) according to the manufacturer’s protocol and as described above. Following 48 h of transfection, media was changed and after 72 h cells were harvested for RNA isolation using TRIzol (Invitrogen). The siRNA (PINK1-siRNA-1, Table S1) targeting PINK1 was a prevalidated siRNA from Ambion (ID #1199) and thus designed and chemically synthesized by Ambion. The control siRNA was predesigned by Ambion to not target any gene in the genome (Silencer®Negative Control#1, Ambion).
Western blot analysis
Cells were washed twice in ice-cold PBS and harvested in a lysis buffer containing 62.5 mM Tris-HCl (pH 6.8), 2% (w/v) SDS and 10% (v/v) glycerol. Protein concentration was determined using the Lowry method, and 20 µg proteins were separated on 12% polyacrylamide gels and transferred on to a PVDF membrane (Amersham Biosciences, Piscataway, NJ, USA) in 48 mM Tris/HCl, 39 mM glycine, 0.037% (w/v) SDS and 15% (v/v) methanol, using a semidry electrophoretic transfer cell (Bio-Rad, Sundbyberg, Sweden). The membrane was blocked in 5% milk for 1 h at room temperature and probed with indicated antibodies overnight at 4°C. The immunoblots were visualized with horseradish peroxidase-conjugated secondary antibodies and enhanced chemiluminescence on Hyperfilm-ECL (Amersham Biosciences). Following immunoblotting, the membrane was stained with Amido Black and a highly abundant band of 
45 kDa was used for normalization. The band intensities were quantified using ImageQuant software. Antibodies used were: COXI: Anti-OxPhos Complex IV subunit I, mouse IgG2a, monoclonal 1D6 (anticytochrome oxidase subunit I) (Molecular Probes, Eugene, OR, USA, Catalog#A-6403), dilution 1:1000; Parkin: Anti Parkin rabbit polyclonal antibody (Cell Signaling, Cat#4211), dilution 1:1000; TFAM: Rabbit polyclonal (52)
, kindly provided by Dr. Claes Gustafsson, Karolinska Institutet, Stockholm, Sweden (Dilution 1:1000).
Glucose uptake assay
Following 72 h of PINK1 siRNA transfection of neuroblastoma cells, glucose uptake was measured. Cells were serum-starved for 2 h and then washed twice with PBS. 3H-2-deoxyglucose was added to glucose-free DMEM and applied to the cells. Following 15 min of incubation, cells were washed twice with cold PBS, subsequently lysed with 500 µl 0.2 M NaOH, and incubated for 30 min at 60°C. 400 µl were used for the scintillation vials and the rest (100 µl) was used for determine protein concentration using the Lowry method. In each experiment, each treatment was determined in triplicate and normalized to the protein concentration.
RNA isolation and quantitative real- time PCR (qRT-PCR)
Human tissue biopsies were homogenized in TRIzol (Invitrogen) using a motor-driven homogenizer (Polytron, Kinematica, Newark, NY, USA) and total RNA was isolated according to the manufacturer’s protocol. Total RNA was dissolved in RNase-free water and quantified using a Spectrophotometer (Pharmacia Biotech, Piscataway, NJ, USA). For the cell samples, total RNA was isolated using TRIzol (Invitrogen), according to the manufacturer’s protocol. Total RNA was reverse-transcribed using reverse transcription reagents (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s protocol. Random hexamers were used for first-strand cDNA synthesis. Detection of mRNA was performed using an ABI-PRISM® 7000 Sequence Detection system (Applied Biosystems). Primers and MGB probes (Table 1)
were designed using Primer Express software (Applied Biosystems) or obtained using the Universal Probe Library (Roche Applied Science). A preoptimized primer and probe assay for 18S rRNA was used as an endogenous control (Applied Biosystems). Primers and probes were premixed with TaqMan Universal Master Mix or SYBR®GREEN PCR Master Mix (Applied Biosystems) and distributed into 96-well MicroAmp Optical barcode plates (Applied Biosystems). cDNA aliquots of 4 µl were added in triplicates. The amplification of genomic DNA typically amounted to a maximum of <1% of the target gene when using the TRIzol protocol. Two-fold dilutions series were performed for all target genes and endogenous controls to determine the amplification efficiency.
Calculations and statistical analysis
All data are presented as mean ± SE, unless otherwise stated. The 
Ct method (Applied Biosystems, User bulletin 2) was used to calculate relative changes in mRNA abundance. The threshold cycle (CT) value for 18S was subtracted from the CT value for the target gene to adjust for any variations in the cDNA synthesis. For the metabolic cohort, a mean was calculated from 18S adjusted CT-values of any target gene for each of the four groups and the individual values in each group were related to that mean. For the paired muscle samples, the preinactivity values reflect baseline gene expression levels and were subtracted from the postinactivity sample. For the siRNA experiments, 18S adjusted CT-values were related to the adjusted CT-value of the untreated sample in each experiment, and then a mean from all experiments was calculated. The results are presented as percentage mRNA abundance of untreated samples (mean±SE). ANOVA was utilized to analyze both the human gene expression responses and the cell based responses. When a significant F ratio was achieved, post hoc analysis (Tukey) was utilized to make individual comparisons and generate P-values that are stated in the figure legends. Sample size and significance level is shown in the figure legends for each graph.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and marker mitochondrial genes in human muscle and adipose tissue
A and C, both the mitochondrial complex I gene, mtND4, and Citrate synthase (CS) were significantly down-regulated in the adipose tissue of obese, normal glucose tolerant subjects (NGT BMI>35) and nonobese type 2 diabetes patients (DM BMI<27) when compared with nonobese, normal glucose tolerant age- and gender-matched controls (NGT BMI<27). Intriguingly, in the obese diabetes group (DM BMI>35), the mitochondrial gene expression was not reduced. In fact, for Citrate synthase there was a significant difference between the nonobese and the obese, type 2 diabetes patients (Fig. 1C
). PGC-1 was down-regulated in all three disease groups in adipose tissue (Fig. 2
A), clearly contrasting with the mitochondrial gene expression profile. In skeletal muscle, only the DM BMI < 27 group demonstrated a reduction in mitochondrial gene expression (Fig. 1B, D
), consistent with the skeletal muscle PGC-1 profile (Fig. 2B
). Noteworthy, 29% of the DM BMI > 35 group were newly diagnosed for type 2 diabetes at the time of the investigation while 100% of the DM BMI < 27 group had preexisting diabetes (Table 1)
. However, when excluding this 29% of DM BMI > 35 individuals from our expression analysis, no impact was noted on the average expression values for PGC-1, Citrate synthase and mtND4 (data not shown). Further analysis demonstrated that expression of PGC-1 in adipose tissue weakly related to aerobic capacity (R2=0.15, Fig. 2C
), while in skeletal muscle there was no correlation (Fig. 2D
). Neither did PGC-1 correlate to self-reported physical activity (data not shown).
|
|
The PINK1 locus genes were reciprocally regulated in skeletal muscle by physical inactivity yet were all suppressed of type 2 diabetes patients
Following on from our custom array analysis, which demonstrated a reduction in muscle PINK1 expression due to inactivity (Supplementary Dataset 1); we studied three main transcripts produced from the PINK1 locus (Fig. 3
). We used qRT-PCR and confirmed that PINK1 was 40% reduced following 5 wk of inactivity in our healthy volunteers (5)
(Fig. 4
A). In contrast, natural antisense PINK1 (naPINK1), a cis-encoded natural antisense transcribed from the PINK1 locus (19)
, tended to be up-regulated in connection with inactivity (Fig. 4A
). This reciprocal pattern of regulation is supported by earlier analysis from an endurance training study (30)
where PINK1 was significantly up-regulated following endurance training, while naPINK1 was significantly down-regulated (19)
. In contrast, DDOST expression was not altered by 5 wk of muscle inactivity (or by 6 wk of endurance training) (Fig. S1). We then sought to establish whether transcript abundance derived from the PINK1 locus related to metabolic fitness. In Fig. 4B-D
, it can be observed that all three transcripts from the locus were down-regulated in diabetes skeletal muscle, providing a clear contrast to the response of this genomic locus to altered physical activity levels. In the NGT BMI > 37 group, PINK1 was not altered while both naPINK1 and DDOST expression were significantly reduced.
|
PINK1 was coexpressed with mitochondrial genes but not PGC-1 in adipose tissue
In contrast to the case in skeletal muscle, transcripts from the PINK1 locus were not significantly altered by obesity or diabetes in adipose tissue (Fig. 5
A–C), although arguably there was some trend for reduced expression. However, PINK1 expression did significantly correlate with the mitochondrially encoded OXPHOS gene mtND4 (R2=0.27, P<0.001) (Fig. 5D
) and the nuclear-encoded mitochondrial gene Citrate synthase (R2=0.185, P=0.0045) (data not shown) in adipose tissue. The relation between PGC-1 and adipose mitochondrial gene expression was weaker (R2=0.11–0.18, P<0.05). Interestingly, there was no correlation between PINK1 and PGC-1 expression in adipose tissue (R2=0.0011, P=0.83, Fig. S2). However, in skeletal muscle, expression of PINK1 and PGC-1 did correlate (R2=0.350, P<0.0001, Fig. S2), presumably reflecting the fact that both genes are regulated by physical activity—a dominant factor for regulating gene expression in skeletal muscle (30)
.
|
PINK1 was not essential for mitochondrial gene expression in a murine brown adipocyte model
To further investigate the potential role of PINK1 in the regulation of mitochondrial gene expression, we utilized a primary brown preadipocyte model (53)
, which demonstrates robust expression of mitochondrial genes. After transfection of short interfering RNA (siRNA) on day 1 of culture, Pink1 was knocked down by 80% on cell harvesting for gene expression 3 days later (Fig. 5)
. In the present model, neither the mitochondria-encoded mtNd5 nor the nuclear encoded Citrate synthase gene expression changed following Pink1 knockdown. At the protein level, we analyzed the expression of the established mitochondrial markers TFAM and COXI, as well as Parkin (a gene suggested to be part of a Pink1 mitochondrial biogenesis pathway in Drosophila). However, none of these proteins showed any alterations in abundance following Pink1 knockdown in our model (Fig. 6
).
|
PINK1 expression related to diabetes status and influenced human cell glucose uptake
Altered expression from the PINK1 locus in skeletal muscle related to a number of diabetes-related parameters in our clinical samples. For example, there was a nonlinear relationship between both HbA1c and fasting glucose levels and both PINK1 and DDOST mRNA levels, where a low expression of either gene was associated with elevated blood glucose levels or HbA1c (Fig. 7
A, B and Fig. S3). It may be noted that in each case a "threshold" type relationship existed, much like recent analysis from the Kahn Laboratory (54)
, where above a certain level for the clinical parameter, gene expression appears most dramatically altered. Surprisingly and in contrast to PINK1, PGC-1 only weakly related to parameters reflecting glucose tolerance (R2 <0.1, Fig. S3). To examine whether the suppression of PINK1 may directly contribute to altered glucose metabolism, we knocked down PINK1 by transfecting siRNAs into human neuroblastoma cells and found that 80% knockdown of PINK1 mRNA for 48 h caused a subsequent reduction of basal glucose uptake in neuroblastoma cells (Fig. 7C-D
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and type 2 diabetes.
PGC-1 and mitochondrial genes were not consistently reduced in skeletal muscle of type 2 diabetes
We have previously shown that PGC-1 is down-regulated in skeletal muscle following inactivity and that several mitochondrial target genes correlated with this down-regulation (5)
. In diabetes patients, loss of PGC-1 is claimed to modulate mitochondrial genes in skeletal muscle (2
, 29)
. Mootha et al. 2003, describe a decrease in the expression of mitochondrial genes in diabetics and a subgroup of these to be regulated by PGC-1. However, the control group in that study had a BMI of 23.6 ± 3.4 (kg/m2), the type 2 diabetes group had a BMI of 27.3 ± 4 (kg/m2) and thus nonobese normal glucose tolerant controls were compared with a blend of nonobese and obese type 2 diabetes subjects. In our study, we distinguish between nonobese (BMI 25.8±1.6) type 2 diabetes subjects and demonstrate a significant down-regulation of PGC-1 and two representative mitochondrial genes (Citrate synthase and mtND4) only in the DM BMI < 27 group. Further, if we excluded the recently-diagnosed patients, this did not alter the mean expression value in the DM BMI > 35 group, such that our observations may reflect the occurrence of a BMI related mechanism that appears to maintain mitochondrial function in type 2 diabetes. Preliminary analysis (Table 1)
suggested this was not due to altered adiponectin levels, which could have been one potential mechanism to explain our observation (55)
, as adiponectin can positively modulate mitochondrial gene expression. Alternatively, obese individuals may develop diabetes by a distinct mechanism, independent of any mitochondrial contribution, while nonobese people with diabetes may develop diabetes for reasons that include impaired mitochondrial function. An attractive feature of this hypothesis is that the confounding effects of inactivity (on molecular data) should have resulted in lower mitochondrial gene expression regardless of the cause-effect debate. This is because it would be expected that obese people are generally less active (an assumption implied but not significantly confirmed by our physical activity questionnaire). While our results are in line with other studies demonstrating a decrease in some but not all mitochondrial OXPHOS genes profiled in skeletal muscle from diabetes (56
57
58)
, it partly challenges the idea proposed by Patti et al. that a decrease in PGC-1 is a specific and early alteration connected with insulin resistance. The patients in that particular study were not remotely age-matched (the control group was 31.1±2.0 yr; insulin resistant, normal glucose tolerant group 40.3±3.0 yr; type 2 diabetes group 47.3±1.9 yr) such that we feel it is difficult to accept the validity of their study. Furthermore, the lack of association between PGC-1 and aerobic capacity or diabetes parameters in our data suggest that PGC-1 can not yet confidently be classified as being an important gene for skeletal muscle dysfunction in diabetes, while genetic associations perhaps reflect the role of PGC-1 in other critical organs such as the liver and the pancreas (59)
.
Altered regulation of the PINK1 locus: a link between type 2 diabetes and neurodegenerative disease?
In line with recent findings of mitochondria and muscle deficient phenotypes in Drosophila Pink1 mutants (14
15
16)
, our data reveal that PINK1, similar to PGC-1, is regulated by physical inactivity in humans. Inactivity is known to be a risk factor for developing several degenerative diseases (60
, 61)
, and as PINK1 is down-regulated by physical inactivity, it may represent one starting point for PINK1-related metabolic dysfunction. The relationship between PINK1 gene expression and fasting glucose levels in human muscle, together with the impaired glucose uptake due to PINK1 knockdown in neuroblastoma cells, implies a potential role for PINK1 in glucose metabolism. We speculate that chronic physical inactivity could in time result in an inactivation of the whole PINK1 locus and that this may subsequently contribute to the development of type 2 diabetes. This could occur via a connection between hyperglycaemia (caused by inactivity induced insulin resistance) and regulation of the PINK1 locus. The different regulation patterns of the PINK1 locus in skeletal muscle during physical inactivity compared to diabetes, indicates that the PINK1 locus may be less available for transcriptional activation in people with diabetes, perhaps due to epigenetic changes (62)
. This kind of genomic modulation has been suggested to play a role in complex diseases such as aging and cancer (63)
, however it is has not yet shown to occur in type 2 diabetes, a research idea that deserves further attention.
Given the linkage between PINK1 mutations and Parkinson’s disease (7)
and the suggestion that Parkinson’s disease patients may have impaired glucose tolerance (11)
, there is a possibility that a decrease in PINK1 may represent a direct link between neurodegeneration and type 2 diabetes, albeit one operating via physical inactivity and skeletal muscle insulin resistance. DDOST, on the other hand, remained unaltered by physical activity. Thus, any altered DDOST expression found in obesity or diabetes is unlikely to directly reflect physical activity levels but rather the subsequent hyperglycemia. The accumulation of AGE in these complex diseases is enhanced by hyperglycemia, and our data suggest that one of the molecular mechanisms through which this may occur could be via down-regulation of DDOST (encoding the AGE-R1 scavenger receptor thought to regulate the removal of AGEs). The previous linkage to Parkinson’s disease implied the assumption that PINK1 plays a causative role in mitochondrial disease. We further extend this idea and suggest that deregulation of the whole PINK1 locus may be involved in the pathophysiology of both neurodegenerative disease and type 2 diabetes.
Implications for a PGC-1 independent pathway for regulation of mitochondrial genes in adipose tissue
Although extensively explored in skeletal muscle (2
, 29
, 56
, 57)
, fewer studies have focused on the gene expression in adipose tissue from people with diabetes. Carey et al, report a discordant regulation in skeletal muscle and adipose tissue of nonobese patients with type 2 diabetes, demonstrating a three-fold up-regulation of PGC-1 and other metabolic genes in adipose tissue, suggesting a potential transformation of white adipose tissue toward a brown adipocyte phenotype using UCP1 gene expression as the sole marker (28)
. We did not observe an induction of PGC-1, rather our results are in line with two previous studies that demonstrated a decrease in PGC-1 expression in adipose tissue of obese subjects and insulin resistant subjects respectively (64
, 65)
while reliance on UCP1 as the sole marker of a brown adipocyte phenotype is not ideal (66)
. Surprisingly, in contrast to PGC-1, Citrate synthase and mtND4 expression were not decreased in the DM BMI > 35 group. This was intriguing as PGC-1 is thought to be a master regulator of mitochondrial genes (59)
, and suggests that a PGC-1 independent pathway operates in human white adipose tissue. This idea was further emphasized by the weak correlation between PGC-1 and the mitochondrial genes in adipose tissue.
Knockdown of Pink1 in murine brown preadipocytes argues against a role in mitochondrial biogenesis
For two reasons we measured the abundance of Parkin following Pink1 knockdown: (1)
Over-expression of Parkin could rescue Drosophila Pink1 loss-of-function mitochondrial disease phenotypes (14
15
16)
but not vice versa (16)
and Pink1 knockdown in Drosophila resulted in a reduced expression of Parkin (14)
, suggesting that Parkin operates down-stream of PINK1 to regulate mitochondrial function (2)
. Parkin has recently been suggested to enhance mitochondrial transcription in mammalian cells (67)
. Interestingly, the expression profile of PINK1 in human adipose tissue also varied closely with mitochondrial gene expression. Our RNAi experiments in murine brown preadipocytes did not alter the expression of mitochondrial proteins at a level of knockdown that had functional consequences for glucose metabolism in neuroblastoma cells. This argues against a role of PINK1 in mitochondrial biogenesis in mammalian adipocytes. More likely, PINK1 has a role in cell energetics, where depletion would not compromise mitochondria in a catastrophic manner, but would rather have a cumulative effect over time, possibly relating to ROS related damage. This hypothesis is in line with a previous study in Drosophila (15)
, where the mitochondria of the Pink1 mutant pupae were unaffected while a degenerative mitochondrial phenotype was observed in the adult Pink1 mutant fly.
In conclusion, our data demonstrate that the PINK1 locus is regulated by physical inactivity, obesity and type 2 diabetes, appears to have a role in glucose metabolism (an observation that require further study in adipocytes and myocytes) but is not apparently essential for mitochondrial biogenesis in brown adipocytes. Thus, we provide one of the first potential molecular links between type 2 diabetes and Parkinson’s disease. We suggest that altered expression of PGC-1 is less likely to have a direct causative role in the skeletal muscle defects observed in diabetes patients, while we demonstrate that down-regulation of PGC-1 and mitochondrial genes in skeletal muscle occurs predominantly in nonobese diabetics, possibly implying an obesity-dependent feedback mechanism for maintaining mitochondrial function or a unique disease pathway for obese subjects becoming diabetic. Future studies should explore the role of PINK1 in glucose metabolism as well as potential epigenetic mechanisms that could explain the silencing of the PINK1 locus and hence the association between type 2 diabetes and neurodegeneration.
|
ACKNOWLEDGMENTS |
|---|
Received for publication March 13, 2007. Accepted for publication May 3, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. Giovannone, W. G. Tsiaras, S. de la Monte, J. Klysik, C. Lautier, G. Karashchuk, S. Goldwurm, and R. J. Smith GIGYF2 gene disruption in mice results in neurodegeneration and altered insulin-like growth factor signaling Hum. Mol. Genet., December 1, 2009; 18(23): 4629 - 4639. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. W. Franks, C. Scheele, R. J. F. Loos, A. R. Nielsen, F. M. Finucane, C. Wahlestedt, B. K. Pedersen, N. J. Wareham, and J. A. Timmons Genomic variants at the PINK1 locus are associated with transcript abundance and plasma nonesterified fatty acid concentrations in European whites FASEB J, September 1, 2008; 22(9): 3135 - 3145. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
