请帮助翻译一篇英文文献

Hum Genet (2008) 124:263–270
DOI 10.1007/s00439-008-0552-2
123
ORIGINAL INVESTIGATIONS
Association of PARL rs3732581 genetic variant with insulin levels,
metabolic syndrome and coronary artery disease
Brenda L. Powell · Steven Wiltshire · Gillian Arscott · Pamela A. McCaskie ·
Joseph Hung · Brendan M. McQuillan · Peter L. Thompson · Kim W. Carter ·
Lyle J. Palmer · John P. Beilby
Received: 17 June 2008 / Accepted: 20 August 2008 / Published online: 30 August 2008
© Springer-Verlag 2008
Abstract PARL (presenilin-associated rhomboid-like) is
a mitochondrial protein involved in mitochondrial membrane
remodelling, and maps to a quantitative trait locus
(3q27) associated with metabolic traits. Recently the
rs3732581 (Leu262Val) variant was found to be associated
with increased levels of plasma insulin, a Wnding not replicated
in a larger cohort. The aim of the current study was to
investigate the associations between rs3732581 and levels
of plasma insulin, metabolic syndrome (MetS) and its components,
and cardiovascular disease. The CUPID population
consisted of 556 subjects with angiographically proven
CAD and the CUDAS cohort consisted of 1,109 randomly
selected individuals from Perth, Western Australia. Samples
were genotyped using mutation-speciWc PCR. No signiWcant
associations were observed between rs3732581 and
levels of plasma insulin, glucose, BMI or MetS in either
population. However, carriers of the minor allele had signiWcantly
lower mean intima-media thickness (IMT)
[0.69 mm, 95% CI (0.69, 0.70 mm); P = 0.004], compared
with major allele homozygotes [mean IMT = 0.71 mm,
95% CI (0.70, 0.72 mm)] in the CUDAS population. Further
analysis using a recessive model showed homozygous
carriers of the minor allele were predisposed to CAD [OR
1.55, 95% CI (1.11, 2.16); P = 0.01]. Despite the functional
evidence for a role of PARL in regulating insulin levels, no
association with rs3732581 was found in the current study.
Additionally, there were no associations with glucose levels,
BMI or MetS. There were signiWcant eVects of the variant
on mean IMT and risk of CAD. A role for PARL in
metabolic conditions cannot be excluded and more comprehensive
genetic studies are warranted.
Introduction
Mitochondrial dysfunction is increasingly being recognised
as contributing to a range of adverse metabolic conditions,
including obesity, insulin resistance (IR) and type 2 diabetes
mellitus (type 2 DM) (Rolo and Palmeira 2006), which
are becoming more prevalent in the developed world.
Recently it has been reported that an age related decrease in
mitochondrial function in skeletal muscle can contribute to
B. L. Powell · S. Wiltshire · K. W. Carter
Western Australian Institute for Medical Research and the UWA
Centre for Medical Research, University of Western Australia,
Perth, Australia
e-mail: bpperth@hotmail.com
G. Arscott · J. P. Beilby (&)
Clinical Biochemistry, PathWest Laboratory Medicine,
J Block, QE II Medical Centre, Hospital Avenue, Nedlands,
Perth, WA 6009, Australia
e-mail: john.beilby@health.wa.gov.au
P. A. McCaskie · L. J. Palmer
Centre for Genetic Epidemiology and Biostatistics,
University of Western Australia, Perth, Australia
P. A. McCaskie
School of Mathematics and Statistics,
University of Western Australia, Perth, Australia
J. Hung · B. M. McQuillan · P. L. Thompson
School of Medicine and Pharmacology,
University of Western Australia, Perth, Australia
J. Hung · B. M. McQuillan · P. L. Thompson
Sir Charles Gairdner Hospital Campus of the Heart Research
Institute of Western Australia, Perth, Australia
J. P. Beilby
School of Surgery and Pathology,
University of Western Australia, Perth, Australia
264 Hum Genet (2008) 124:263–270
123
IR in elderly people (Petersen et al. 2003). Skeletal muscle
is the major site of peripheral IR and defects in insulin signalling
in the muscle can contribute to increased levels of
glucose in the blood (Reaven 1995).
A mitochondrial protein, Presenilin-associated rhomboid-
like (PARL), that interacts with the proteins associated
with familial Alzheimer’s disease, presenilin-1 and
presenilin-2 has been identiWed (Pellegrini et al. 2001).
Endogenous PARL is constitutively phosphorylated, inhibiting
mitochondrial fragmentation that is a precursor to cellular
apoptosis (Jeyaraju et al. 2006). Recently, PARL has
been reported to be diVerentially expressed in skeletal muscle
of rats (Psammomys obesus) symptomatic for type 2
DM, compared with rats of normal glucose tolerance. Exercise
training, ameliorating symptoms of type 2 DM by
reducing the plasma levels of glucose and insulin, also
increased the levels of PARL in skeletal muscle (Walder
et al. 2005). Analysis of PARL expression in human subjects
conWrmed this Wnding, but only in subjects without a
family history of type 2 DM. In one of only two genetic
studies carried out on PARL to date, a common genetic variant
(Leu262Val; rs3732581; c.843C>G) was genotyped in
individuals of Northern European ancestry where it was
found to be associated with increased levels of fasting
plasma insulin (Walder et al. 2005). This Wnding was not
replicated in a larger study by Fawcett et al. (2006).
The chromosomal location of PARL is on 3q27, a region
reportedly linked to phenotypes associated with metabolic
syndrome (MetS) including BMI and insulin levels (Kissebah
et al. 2000). Due to the previous conXicting association
results of the PARL rs3732581 variant with plasma insulin
levels, the aim of the current study was to investigate the
relationship between the rs3732581 variant and fasting
plasma insulin levels in two well-characterised Western
Australian populations. In addition, associations between
this genetic variant and fasting plasma glucose levels and
BMI were also investigated. With the strong association of
IR and adiposity with MetS, we investigated the eVect of
PARL rs3732581 with respect to the MetS. Finally, as mitochondrial
dysfunction has been reported to be associated
with the rupture of atherosclerotic plaque (Mallat and Tedgui
2000), the eVect of this variant on cardiovascular disease
was explored.
Materials and methods
Subjects
This study examined participants in two Western Australian
cohorts—the Carotid Ultrasound Disease Assessment
Study (CUDAS) and the Carotid Ultrasound in Patients
with Ischaemic Heart Disease (CUPID) study (Chapman
et al. 2001; McCaskie et al. 2006a, b; McQuillan et al.
1999). Both the populations were recruited in the mid
1990s and are predominantly European-Australian. The
CUDAS population (n = 1,109) consists of subjects who
were original participants in the 1989 Australian National
Heart Foundation (NHF) Perth Risk Factor Prevalence Survey.
This was a random electoral roll survey of 2,000 people
from the metropolitan area of Perth, with equal numbers
of males and females and equal numbers of subjects in each
age decile between 20 and 70 years. Patients in the CUPID
cohort (n = 556) are predominantly male and aged between
26 and 60 years. All patients had a history of angina, unstable
angina or myocardial infarction (MI) and angiographically
proven coronary artery disease (CAD) with at least
one coronary vessel with >50% stenosis, but were otherwise
medically stable. In addition, we created a case/control
sample comprising the 485 males from CUPID together
with the 502 males from CUDAS. Written, informed consent
was obtained from all study participants. The study
protocol was approved by the Human Research Ethics
Committee at the University of Western Australia.
Laboratory measures
A fasting venous blood sample was obtained from each
subject. Total cholesterol (TC), HDL cholesterol (HDL-C),
low-density lipoprotein cholesterol (LDL-C) and triglyceride
(TG) levels were determined as previously described
(Chapman et al. 2001). Fasting plasma insulin and glucose
levels were measured, and IR was calculated using the
homeostatic model assessment score [HOMA; fasting
plasma insulin (mU/L) £ fasting plasma glucose (mmol/L)/
22.5] (Matthews et al. 1985).
Clinical measures
Focal plaque and mean intima-media thickness (IMT) were
determined as previously described (Chapman et al. 2001;
McQuillan et al. 1999). Disease severity in CUPID patients
was assessed using the Gensini score (Gensini 1983). The
Gensini score system yields a qualitative and quantitative
evaluation of the coronary angiogram which grades narrowing
of the lumen of the coronary artery. The score is
then multiplied by a factor that takes into account the
importance of the position of the lesion in the coronary
arterial tree. The same experienced angiographer reviewed
all the angiograms independently in order to measure
Gensini score. Patients were asked to report a history of
smoking and physician-diagnosed hypertension, diabetes,
hyperlipidemia, angina pectoris, myocardial infarction, or
stroke, and to include medication use. Blood pressure and
anthropomorphic measures were recorded by a trained
nurse.
Hum Genet (2008) 124:263–270 265
123
Metabolic syndrome
Patients were classiWed as having MetS based on both International
Diabetes Federation (IDF) and National Cholesterol
Education Program (NCEP) Adult Treatment Panel III
(ATPIII) criteria. IDF criteria requires the presence of central
obesity (deWned as waist circumference ¸94 cm for
Europid men and ¸80 cm for Europid women) plus any two
of the following four factors—raised TG level: ¸150 mg/dL
(1.7 mmol/L), or speciWc treatment for this lipid abnormality;
reduced HDL-C: <40 mg/dL (1.03 mmol/L) in males
and <50 mg/dL (1.29 mmol/L) in females, or speciWc treatment
for this lipid abnormality; raised blood pressure: systolic
BP ¸130 or diastolic BP ¸85 mmHg, or treatment of
previously diagnosed hypertension; and raised fasting
plasma glucose (FPG) ¸100 mg/dL (5.6 mmol/L), or previously
diagnosed type 2 DM (Zimmet et al. 2005). NCEP
ATPIII criteria requires at least three of the following—central
obesity, waist circumference ¸102 cm (male), ¸88 cm
(female); TG ¸1.7 mmol/L (150 mg/dl), or drug treatment
for elevated TG; HDL-C <40 mg/dL (1.03 mmol/L) in
males and <50 mg/dL (1.29 mmol/L) in females, or drug
treatment for reduced HDL-C; systolic BP ¸130 or diastolic
BP ¸85 mm Hg, or drug treatment for hypertension; and
fasting plasma glucose ¸100 mg/dL (5.6 mmol/L), or drug
treatment for elevated glucose (Grundy et al. 2005).
Genotyping
Genomic DNA was extracted from leucocytes using a salting
out method. PARL rs3732581 was analysed by mutation-
speciWc PCR. Genotyping was performed by the
PathWest Molecular Genetics Service, Western Australia.
All primer sequences and full experimental conditions are
available upon request.
Statistical analyses
Primary quantitative outcomes for the association analyses
were fasting plasma insulin and glucose measures, and BMI.
The primary binary outcome was MetS, diagnosed according
to both NCEP ATPIII and IDF criteria. The principal explanatory
variable was the PARL rs3732581 single nucleotide polymorphism
(SNP). Genotypes were examined for deviation
from Hardy–Weinberg equilibrium (HWE) using a chi-square
goodness-of-Wt test. The SNP genotypes were coded into three
classes and analysed categorically with the most common
homozygous genotype as a reference category. Genetic eVect
types were determined to be codominant, dominant or recessive
by analysing the trend of the beta-coeYcients for each
category. If a dominant or recessive model was suggested by
these coeYcients to better Wt the data than under a codominant
assumption, the relevant model was then performed.
To investigate the impact of the variant on the risk of
developing cardiovascular disease, we explored four related
outcomes. In the CUDAS population, the eVect of PARL
rs3732581 on mean IMT and presence of plaque was determined.
A male only case-control study was formed through
selection of 485 male subjects from CUPID and 502 male
subjects with a similar age distribution (without prior evidence
of cardiovascular disease) from the CUDAS population.
The diVerence in distribution of the variant within this
sample, with respect to CAD, was investigated. Finally, the
eVect of the variant on the severity of disease using Gensini
score (CUPID only), was analysed.
Gender, physician-diagnosed diabetes, use of antihypertensive
and cholesterol-lowering medications and MetS
were analysed as binary variables. All other variables,
including pack years of cigarette smoking, were analysed as
continuous variables. Insulin, glucose and triglyceride levels,
as well as mean IMT measures, were log-transformed
prior to statistical analysis to achieve Normality. Diabetic
subjects were excluded from analyses of insulin and glucose
in the CUPID and CUDAS samples.
Independent predictors of binary and quantitative outcome
variables were determined via a stepwise variable
selection. Generalised linear models (McCullagh and
Nelder 1989) were used to model the eVects of multiple
covariates on binary and quantitative response variables.
Statistical analyses were carried out using the SimHap
v1.0.0 software package (McCaskie et al. 2006a, b). As our
study involves testing multiple related phenotypes, within
two separate hypotheses—the healthy state and the atherosclerotic
disease state—we consider a Bonferroni correction
to be conservative. Consequently, results were deemed
nominally signiWcant at the 5% level, and are presented
unadjusted for multiple testing, for interpretation within the
context of the study.
Power
Power calculations were performed using the PS Power Calculator
(Dupont and Plummer 1997), and the Genetic Power
Calculator (Purcell et al. 2003). The CUDAS sample had
78% power to detect a diVerence in genotypic means for a
quantitative trait of 0.2SD for PARL rs3732581 under a
recessive model, and 85% under a dominant model. Further,
this study had approximately 80% power to detect diVerence
in OR of 1.25 between genotypes for binary traits, calculated
in both the CUDAS and case-control populations.
Results
Table 1 details clinical characteristics of the study populations.
The CUPID cohort contained signiWcantly more
266 Hum Genet (2008) 124:263–270
123
males (87%) than the CUDAS sample (50%), reXecting the
greater prevalence of CAD amongst males. The CUPID
population had lower levels of LDL cholesterol, probably
due to the higher use of cholesterol-lowering medications
in CUPID. Mean levels of fasting plasma insulin were
higher in CUPID, which is probably the result of higher
measures of obesity (BMI, waist circumference) reported
for this group. In CUPID, more participants were diagnosed
as having MetS, a history of MI, and especially diabetes
(pertinent given the proposed involvement of PARL), than
in CUDAS; this reXects the strong relationship of these
parameters with CAD. Additionally, CUPID subjects were
also more likely to be on current cholesterol-lowering and
antihypertensive medications. Genotype frequencies for
both the CUDAS and CUPID populations, as well as the
case-control subset, are detailed in Table 2. Genotypic distribution
was consistent with HWE in both the populations.
We examined the eVect of variation at rs3732581 on the
levels of fasting plasma insulin and glucose, as well as on
BMI in the CUDAS and CUPID cohorts (Table 3). After
adjusting for conventional risk factors, no signiWcant diVerence
with respect to genotype was observed in CUDAS for
insulin (P = 0.76), glucose (P = 0.20) or BMI (P = 0.98).
When these analyses were performed in CUPID, no signiWcant
associations were observed (insulin, P = 0.19; glucose,
P = 0.60; BMI, P = 0.93).
An association between the PARL rs3732581 variant and
MetS (adjusted for independent predictors of MetS) was
investigated in the CUDAS and CUPID samples (Table 3).
No association was observed in either CUDAS (NCEP
ATPIII diagnosis, P = 0.83; IDF diagnosis, P = 0.46) or
CUPID (NCEP ATPIII diagnosis, P = 0.50; IDF diagnosis,
P = 0.93).
When investigating the association between preclinical
cardiovascular disease (mean IMT and plaque) and genetic
variation at rs3732581, the presence of the minor C allele
was associated with a decrease of 0.02 mm in mean IMT
(P = 0.01, Table 4). This association was further investigated
using a dominant model. Carriers of the minor allele
had signiWcantly lower mean IMT [0.69 mm, 95% CI (0.69,
0.70 mm); P = 0.004], compared with individuals carrying
the wild-type genotype (mean IMT = 0.71 mm, 95% CI
(0.70, 0.72 mm)). No association of the variant with the
presence of plaque was observed (P = 0.55, Table 4).
Subsequent analysis of risk of CAD in the case-control
study revealed an association between the C allele and an
increased risk of CAD (P = 0.04, Table 4). Further analysis
under a recessive model revealed that homozygous carriers
of the minor C allele were more likely to have CAD [OR
1.55 95% CI (1.11,2.16); P = 0.01] compared to other
genotypes. No association with the severity of disease (as
determined by Gensini score in CUPID) was observed
(P = 0.46, Table 4).
Table 1 Description of the study populations
CUDAS Carotid Ultrasound Disease Assessment Study
CUPID Carotid Ultrasound in Patients with Ischaemic Heart Disease
MI myocardial infarction
NCEP National Cholesterol Education Program Adult Treatment
Panel III
IDF International Diabetes Federation
Variable (units) CUDAS (1,109)
(mean and SD)
CUPID (556)
Age (years) 53 (13) 50 (5)
BMI (kg/m2) 26.1 (4.1) 28.3 (4.1)
WHR 0.84 (0.09) 0.92 (0.07)
Waist (cm) 84.8 (12.4) 95.2 (11.4)
Systolic BP (mmHg) 128 (19) 126 (16)
Diastolic BP (mmHg) 80 (10) 81 (10)
Total cholesterol (mmol/L) 5.6 (1.0) 5.2 (1.0)
LDL (mmol/L) 3.6 (0.9) 3.2 (0.9)
Triglycerides (mmol/L) 1.3 (0.7) 2.0 (1.4)
HDL (mmol/L) 1.3 (0.4) 1.1 (0.3)
Glucose (mmol/L) 5.5 (1.3) 5.4 (2.0)
Insulin (pmol/L) 45.2 (41.0) 65.2 (46.6)
Pack years of smoking 12.3 (21.4) 23.1 (25.8)
Mean IMT (mm) 0.71 (0.14) NA
Gensini score NA 10.88 (3.82)
Variable n (%)
Males 50.3 (558) 87.2 (485)
Physician-diagnosed diabetes 2.1 (23) 15.6 (87)
NCEP deWned metabolic syndrome 24.6 (273) 40.0 (220)
IDF deWned metabolic syndrome 26.9 (298) 45.4 (251)
Cholesterol-lowering medication 6.8 (75) 66.7 (371)
Antihypertensive medication 15.5 (172) 52.7 (293)
Presence of plaque 25.6 (284) 59.5 (331)
Table 2 Genotype and allele distribution within CUDAS, CUPID,
and the case-control cohort
Genotype counts include all individuals
PARL CUDAS CUPID Controls Cases
Genotype
GG 312 (28.2) 158 (28.4) 157 (31.3) 131 (27.0)
GC 564 (50.9) 257 (46.2) 252 (50.3) 228 (47.0)
CC 232 (20.9) 141 (25.4) 92 (18.4) 126 (26.0)
Total 1,108 556 501 485
Allele
G 1,188 (53.6) 573 (51.5) 566 (56.5) 490 (50.5)
C 1,028 (46.4) 539 (48.5) 436 (43.5) 480 (49.5)
Total 2,216 1,112 1,002 970
Hum Genet (2008) 124:263–270 267
123
Table 3 Estimated marginal means (EMM) and odds ratios (OR) of insulin, glucose, BMI and metabolic syndrome for the PARL rs3732581 sequence variant within the CUDAS and CUPID
cohorts
Analyses of insulin and glucose levels were carried out in non-diabetic patients only
Models adjusted as below
CUDAS Insulin: age, BMI, gender, glucose, SBP, TG and WHR; Glucose: age, gender and insulin; BMI: gender, HDL, insulin, SBP and WHR; NCEP ATPIII: age, BMI, gender and insulin;
IDF: age, BMI and insulin
CUPID Insulin: BMI, cholesterol medication, glucose and TG; Glucose: age, insulin and LDL; BMI: gender, insulin, SBP and WHR; NCEP ATPIII: BMI and insulin; IDF: BMI, cholesterol
medication, gender and insulin
EMM estimated marginal means
Insulin (pmol/L) Glucose (mmol/L) BMI (kg/m2) Metabolic syndrome
(NCEP ATPIII)
Metabolic
syndrome (IDF)
EMM (95% CI) P value EMM (95% CI) P value EMM (95% CI) P value OR (95% CI) P value OR (95% CI) P value
CUDAS
GG 36.65 (34.62, 38.79) 5.40 (5.30, 5.50) 26.10 (25.76, 26.44) 1.0 1.0
GC 35.80 (34.32, 37.34) 5.32 (5.25, 5.40) 26.07 (25.82, 26.33) 1.03 (0.70, 1.53) 0.81 (0.54, 1.21)
CC 35.62 (33.34, 38.05) 0.76 5.26 (5.15, 5.38) 0.20 26.12 (25.73, 26.51) 0.98 1.15 (0.71, 1.85) 0.83 0.75 (0.46, 1.24) 0.46
CUPID
GG 59.12 (54.24, 64.43) 4.66 (4.53, 4.79) 27.91 (27.38, 28.44) 1.0 1.0
GC 56.96 (53.10, 61.10) 4.73 (4.63, 4.84) 27.83 (27.40, 28.25) 0.92 (0.54, 1.57) 0.95 (0.55, 1.65)
CC 63.04 (57.66, 68.92) 0.19 4.74 (4.61, 4.87) 0.60 27.96 (27.41, 28.52) 0.93 0.71 (0.39, 1.29) 0.50 0.89 (0.49, 1.63) 0.93
268 Hum Genet (2008) 124:263–270
123
Discussion
The importance of mitochondrial processes in metabolic
conditions is now apparent, as defective oxidative pathways
in skeletal muscle obtained from obese patients and those
diagnosed with type 2 DM have been reported (Kim et al.
2000; Petersen et al. 2003). Recently Walder et al. (2005)
reported Wnding a mitochondrial protein, PARL, with
reduced expression in skeletal muscle of rats (Psammomys
obesus)—an animal model having a number of adverse
metabolic phenotypes including obesity, IR, dyslipidaemia
and type 2 DM. When investigated in subjects from an
American cohort with a high mean BMI (33.4 § 7.2 kg/
m2), the PARL rs3732581 genetic variant was associated
with increased levels of fasting plasma insulin. A recent
report using two large population based cohorts from the
UK has thrown doubt on this association (Fawcett et al.
2006). The mean BMI in the UK study (27.3 § 4.5 kg/m2)
was less than that of the American study; however, the
authors accounted for this in their analyses and failed to
replicate the initial Wndings.
The current study investigated the eVect of the PARL
rs3732581 genetic variant, the only reported coding nonsynonymous
variant in a European population, on fasting
plasma insulin and glucose levels, as well as BMI, in a
healthy population and in a population diagnosed with
CAD. We found no signiWcant eVect of the minor C allele
neither on levels of fasting plasma insulin and glucose, nor
with BMI. These Wndings further support the recent work
showing that the PARL rs3732581 variant has no eVect on
these parameters in population-based cohorts (Fawcett
et al. 2006). In addition, the current results show no association
of the PARL rs3732581 variant with these IR phenotypes
in a population diagnosed with coronary disease,
despite the strong functional evidence that PARL is
involved in regulating insulin levels (Walder et al. 2005).
A previous study reported a strong association of the
genomic location of PARL (3q27) with phenotypes typically
associated with MetS (Kissebah et al. 2000). Despite
the general agreement that MetS is characterised by a cluster
of unfavourable metabolic phenotypes, diagnosis of the
condition is not consistent across countries. There are two
major MetS deWnitions currently in use—one is based on a
consensus reached in 2005 by the International Diabetes
Federation (Zimmet et al. 2005) and the other is based on
the guidelines proposed by the National Cholesterol Education
Program (NCEP) Adult Treatment Panel III (ATP III)
in 2001 (Anonymous 2001) and updated in 2005 (Grundy
et al. 2005). Regardless of which deWnition was used, no
association was found between the PARL rs3732581 variant
and MetS, despite this study being well powered to
detect the eVects of modest size.
Recently, it has been suggested that mitochondrial dysfunction
is involved in atherosclerosis, independent of IR
(Semenkovich 2006). To determine if the PARL rs3732581
variant had any role in coronary disease, we carried out an
exploratory analysis, looking at the eVect this variant had
on mean IMT and plaque in healthy individuals, the presence
of CAD in a case-control setting and the severity
(based on Gensini score) of angiographically proven CAD.
As the mode of action of this variant is currently unknown,
the initial analyses were carried out using a codominant
model and subsequently reWned based on the trend of betacoeYcients.
From these results, the variant appeared to be
acting in a protective dominant manner in a healthy population
(P = 0.004), but in a predisposing recessive manner in
Table 4 EVect of PARL rs3732581 on preclinical and clinical cardiovascular disease
Diabetic subjects were not excluded in these analyses
Phenotypes in CUDAS were adjusted as follows
IMT: age, gender, LDL, pack years of smoking, SBP, WHR and age:gender
Plaque: age, diabetes, gender, LDL, pack years of smoking and SBP
Phenotypes in the case/control were adjusted as follows
CAD: age, diabetes, SBP, TG and WHR
Phenotypes in CUPID were adjusted as follows
Gensini: age, cholesterol medication, gender, HDL and SBP
EMM estimated marginal means
Genotype CUDAS Case/control CUPID
EMM (95% CI) P value OR (95% CI) P value OR (95% CI) P value EMM (95% CI) P value
IMT Plaque CAD Gensini
GG 0.71 (0.70, 0.72) 1.0 1.0 11.14 (10.54, 11.74)
GC 0.69 (0.69, 0.70) 0.90 (0.61, 1.31) 1.05 (0.76, 1.45) 10.91 (10.44, 11.38)
CC 0.69 (0.68, 0.70) 0.01 0.77 (0.48, 1.23) 0.55 1.59 (1.08, 2.36) 0.04 10.60 (9.98, 11.22) 0.46
Hum Genet (2008) 124:263–270 269
123
a CAD case-control cohort (P = 0.01). There was no association
of the variant with the presence of plaque in a healthy
population or severity of CAD (codominant model,
P = 0.55, and P = 0.46, respectively).
While the apparently opposing eVects of the PARL
genotype on IMT and CAD risk are diYcult to reconcile,
they may highlight diVerences in the biological pathways
between the two phenotypes, although we cannot rule out
the possibility that one or more of our results represent
false positive Wndings. The thickening of the IMT is a slow
progressive form of atherosclerosis while clinically signiWcant
CAD is due to the rupture of unstable plaques, possibly
due to apoptosis of cells within these plaques (Mallat
and Tedgui 2000). Comparative studies of yeast and mammalian
mitochondrial membrane dynamics and remodelling
have shown that PARL possesses antiapoptotic
activity by controlling the release of cytochrome-c through
its interaction with, and processing of, the GTPase OPA1
complex (McQuibban et al. 2003; Cipolat et al. 2006).
Phosphorylation of PARL results in an inhibition of mitochondrial
fragmentation and ultimately inhibits apoptosis
(Jeyaraju et al. 2006). It is tempting to speculate that altering
the normal activity of PARL could lead to plaques that
are more susceptible to rupture due to an increase in apoptotic
activity. Potentially the same increase in cellular
apoptotic activity may lead to a thinner carotid IMT and
hence reconcile the diVerent eVects of the genotype on
IMT and CAD risk. Nevertheless, both our original Wndings
require replication in additional populations and any
biological consequences such as those speculated upon
here would need functional investigation before the true
impact of this polymorphism on vascular biology is to be
fully appreciated.
It has been reported that increased expression of PARL
is associated with improved insulin sensitivity, in both animal
and human studies (Walder et al. 2005). While it is
likely that PARL rs3732581 was initially chosen for genetic
analysis as it is the only reported coding non-synonymous
polymorphism in Europeans, there have been no studies
published reporting on the functional eVect this SNP has on
PARL expression or metabolism. It is tempting to speculate
that the Leu262Val substitution in PARL could in some
way inXuence the antiapoptotic activity of PARL and mitochondrial
remodelling through its interaction with OPA1
(McQuibban et al. 2003; Cipolat et al. 2006). Apoptosis is
believed to be important in regulating beta-cell mass in the
pancreas, and may contribute to the development of type 2
DM (Lupi and Del Prato 2008). Mitochondrial dysfunction
can lead to impairments of lipid and glucose metabolism
and glucose-stimulated insulin release (Civitarese and
Ravussin 2008). However, our study, like that of Fawcett
et al. and the many recent genome-wide association scans
of type 2 DM, provides no evidence for the Leu262Val
polymorphism having a role in type 2 DM susceptibility or
glucose homeostasis.
The extent of genetic variation in this gene has not been
comprehensively catalogued, therefore we cannot rule out
the possibility that there exists other functional variants,
such as those that could aVect protein phosphorylation,
cleavage or apoptotic function (Jeyaraju et al. 2006). Therefore,
in light of our Wndings with respect to CAD and the
importance of PARL in mitochondrial biology and glucose
metabolism, a comprehensive survey of the genetic variation
in PARL and its eVects on metabolic conditions is warranted.
It is recommended that a thorough genetic
investigation using resequencing or tagging methods
should be performed, and that such an investigation may
reveal variants important in the pathogenesis of adverse
metabolic conditions.
Acknowledgments This study was funded by the National Heart
Foundation grant-in-aid G97P 5002. BLP was supported by a National
Health and Medical Research Council Howard Florey Centenary Fellowship
(Grant ID: 404129). KWC was supported by the Australian
Research Council. The authors thank the participants of the CUPID
and CUDAS cross-sectional studies.
References
Anonymous (2001) Executive summary of the Third Report of the National
Cholesterol Education Program (NCEP) Expert Panel on
Detection, Evaluation, and Treatment of High Blood Cholesterol
in Adults (Adult Treatment Panel III). J Am Med Assoc
285:2486–2497
Chapman CM, Palmer LJ, McQuillan BM, Hung J, Burley J, Hunt C,
Thompson PL, Beilby JP (2001) Polymorphisms in the angiotensinogen
gene are associated with carotid intimal-medial thickening
in females from a community-based population.
Atherosclerosis 159:209–217
Cipolat S, Rudka T, Hartmann D, Costa V, Serneels L, Craessaerts K,
Metzger K, Frezza C, Annaert W, D’Adamio L, Derks C, Dejaegere
T, Pellegrini L, D’Hooge R, Scorrano L, De Strooper B (2006)
Mitochondrial rhomboid PARL regulates cytochrome c release
during apoptosis via OPA1-dependent cristae remodeling. Cell
126:163–175
Civitarese AE, Ravussin E (2008) Minireview: mitochondrial energetics
and insulin resistance. Endocrinology 149:950–954
Dupont WD, Plummer WD (1997) PS power and sample size program
available for free on the Internet. Control Clin Trials 18:274
Fawcett KA, Wareham NJ, Luan J, Syddall H, Cooper C, O’Rahilly S,
Day IN, Sandhu MS, Barroso I (2006) PARL Leu262Val is not
associated with fasting insulin levels in UK populations. Diabetologia
49:2649–2652
Gensini GG (1983) A more meaningful scoring system for determining
the severity of coronary heart disease. Am J Cardiol 51:606
Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin
BA, Gordon DJ, Krauss RM, Savage PJ, Smith SC Jr, Spertus JA,
Costa F (2005) Diagnosis and management of the metabolic syndrome.
An American Heart Association/National Heart, Lung,
and Blood Institute ScientiWc Statement. Executive summary.
Cardiol Rev 13:322–327
Jeyaraju DV, Xu L, Letellier MC, Bandaru S, Zunino R, Berg EA,
McBride HM, Pellegrini L (2006) Phosphorylation and cleavage
270 Hum Genet (2008) 124:263–270
123
of presenilin-associated rhomboid-like protein (PARL) promotes
changes in mitochondrial morphology. Proc Natl Acad Sci USA
103:18562–18567
Kim JY, Hickner RC, Cortright RL, Dohm GL, Houmard JA (2000)
Lipid oxidation is reduced in obese human skeletal muscle. Am J
Physiol Endocrinol Metab 279:E1039–E1044
Kissebah AH, Sonnenberg GE, Myklebust J, Goldstein M, Broman K,
James RG, Marks JA, Krakower GR, Jacob HJ, Weber J, Martin
L, Blangero J, Comuzzie AG (2000) Quantitative trait loci on
chromosomes 3 and 17 inXuence phenotypes of the metabolic
syndrome. Proc Natl Acad Sci USA 97:14478–14483
Lupi R, Del Prato S (2008) Beta-cell apoptosis in type 2 diabetes:
quantitative and functional consequences. Diabetes Metab
34:S56–S64
Mallat Z, Tedgui A (2000) Apoptosis in the vasculature: mechanisms
and functional importance. Br J Pharmacol 130:947–962
Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF,
Turner RC (1985) Homeostasis model assessment: insulin resistance
and beta-cell function from fasting plasma glucose and
insulin concentrations in man. Diabetologia 28:412–419
McCaskie PA, Cadby G, Hung J, McQuillan BM, Chapman CM, Carter
KW, Thompson PL, Palmer LJ, Beilby JP (2006a) The C-480T
hepatic lipase polymorphism is associated with HDL-C but not
with risk of coronary heart disease. Clin Genet 70:114–121
McCaskie PA, Carter KW, Palmer LJ (2006) SimHap: a comprehensive
modelling framework and a simulation-based approach to haplotypic
analysis of population-based data. http://www.genepi.
org.au/simhap. Accessed 31 May 2007
McCullagh P, Nelder J (1989) Generalised linear models. Chapman
and Hall, London
McQuibban GA, Saurya S, Freeman M (2003) Mitochondrial membrane
remodelling regulated by a conserved rhomboid protease.
Nature 423:537–541
McQuillan BM, Beilby JP, Nidorf M, Thompson PL, Hung J (1999)
Hyperhomocysteinemia but not the C677T mutation of methylenetetrahydrofolate
reductase is an independent risk determinant
of carotid wall thickening. The perth carotid ultrasound disease
assessment study (CUDAS). Circulation 99:2383–2388
Pellegrini L, Passer BJ, Canelles M, Lefterov I, Ganjei JK, Fowlkes BJ,
Koonin EV, D’Adamio L (2001) PAMP and PARL, two novel
putative metalloproteases interacting with the COOH-terminus of
Presenilin-1 and -2. J Alzheimers Dis 3:181–190
Petersen KF, Befroy D, Dufour S, Dziura J, Ariyan C, Rothman DL,
DiPietro L, Cline GW, Shulman GI (2003) Mitochondrial dysfunction
in the elderly: possible role in insulin resistance. Science
300:1140–1142
Purcell S, Cherny SS, Sham PC (2003) Genetic power calculator:
design of linkage and association genetic mapping studies of
complex traits. Bioinformatics 19:149–150
Reaven GM (1995) Pathophysiology of insulin resistance in human
disease. Physiol Rev 75:473–486
Rolo AP, Palmeira CM (2006) Diabetes and mitochondrial function:
role of hyperglycemia and oxidative stress. Toxicol Appl Pharmacol
212:167–178
Semenkovich CF (2006) Insulin resistance and atherosclerosis. J Clin
Invest 116:1813–1822
Walder K, Kerr-Bayles L, Civitarese A, Jowett J, Curran J, Elliott K,
Trevaskis J, Bishara N, Zimmet P, Mandarino L, Ravussin E,
Blangero J, Kissebah A, Collier GR (2005) The mitochondrial
rhomboid protease PSARL is a new candidate gene for type 2
diabetes. Diabetologia 48:459–468
Zimmet P, Magliano D, Matsuzawa Y, Alberti G, Shaw J (2005) The
metabolic syndrome: a global public health problem and a new
deWnition. J Atheroscler Thromb 12:295–300